Eligibility criteria

We included primary qualitative studies (and mixed methods studies that include a qualitative component) that document experiences of sexual violence against people in transit, while travelling without authorization. We excluded work that relied only on conceptual development or that only provided quantitative with no qualitative data. We included both peer reviewed academic research and grey literature in the form of doctoral theses and non-governmental organisation reports, as the inclusion of grey literature helps minimise publication bias and ensure a more comprehensive understanding of the available evidence [41]. Additional sources such as book reviews, conference papers and editorials were excluded. We included studies published since 1 January 2010, reporting violence that occurred since 2000, and no restrictions were placed on study setting (country) and source language. No results were returned outside of the language competency of the researchers.

As recommended for qualitative reviews, we followed a Population-Exposure-Outcome (PEO) design. The population were migrants in insecure status, following Innes’ definition [42]. The exposure was an undocumented and unauthorized migration journey, which was defined as being in transit, already outside of the home country, with no status requested or acquired in a destination country at the time of violence occurring. This meant that violence was always a function of the journey. Cases of sexual violence in a destination country, such as intimate partner or family violence following immigration, and cases where the purpose of migration was sexual exploitation in a destination country were not included unless the sexual violence happened during and as a consequence of the journey.

The outcome was sexual violence, which followed the World Health Organization’s definition of sexual violence. In line with this definition, we included verbal, physical, and emotional forms of sexual violence, and threat of violence [43].

Search strategy

The search strategy combined three concept clusters: sexual violence, migration and methods. The concept clusters were developed by the first and second reviewers, with opinion sought from experts in violence and migration studies external to the reviewers. A Boolean search was carried out to comprehensively gather articles citing the separate clusters (OR search) and then combine the clusters with each other (AND search). See S2 Table for the full search strategy.

Database selection was based on previous reviews of violence related to insecure migration status that combined areas of expertise across authorship [2,3]. The five selected databases were Embase, Social Policy and Practice, Political Science Complete, Sociology Source Ultimate, and Web of Science Social Sciences Citation Index and searches were run on 16 January 2025. These database searches were supplemented with searches of grey literature. Grey literature searches were conducted between March and April 2025 using relevant key words in the search function and/or hand searches of research or publication sections of websites of the following organisations: European Commission; Freedom Fund; International Migration, Integration and Social Cohesion Research Network; International Labour Organisation; International Organisation for Migration; La Strada International; International Federation of Red Cross and Red Crescent Societies; Derechos Humanos Fray Matias de Cordova; World Health Organisation; United Nations Office on Drugs and Crime; Walk Free Foundation; SEREDA Project. In addition, the PROQUEST database was used to identify potentially relevant theses and dissertations. All included studies were also subject to forwards and backwards citation tracking to identify potential additional studies for inclusion. Forwards citation tracking was conducted using Google Scholar.

Study selection and data extraction

Zotero open-source referencing software [44] was used to deduplicate the search results, and the systematic review software Rayyan [45] was used for both title and abstract screening and for full-text screening. The first reviewer screened all titles and abstracts against the predetermined inclusion and exclusion criteria and studies that met the inclusion criteria at this stage proceeded to full-text assessment. Out of the 590 records initially screened at title and abstract stage, 211 proceeded to full-text screening, with 205 ultimately screened against the inclusion and exclusion criteria (see S3 Table) at full-text stage (full-texts of 6 records were not retrieved). The second reviewer independently screened 20% of records at both review stages and any disagreements were resolved through discussion until a consensus was reached. Following full-text screening, a total of 52 reports (representing 49 separate studies) met the inclusion criteria and were included in the review (Fig 1). These reports were categorised into two groups: 1) a full-inclusion group subjected to qualitative coding (n = 27) comprising studies with substantial qualitative extracts on sexual violence (3 + mentions), and 2) a route-only group (n = 25) made up of studies that met inclusion criteria but contained limited qualitative extracts on sexual violence (<3 mentions). While the full inclusion group is the main focus of the findings, the route-only group provides geographic context regarding research coverage of sexual violence along migration routes.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. PRISMA Flow Diagram.

https://doi.org/10.1371/journal.pone.0355474.g001

Data collection process

A standardised Excel data extraction template was developed to capture relevant information from included studies. The template recorded the following categories: report ID and year, sexual violence type, source of violence information (e.g., first-hand account, witness etc.), country or region of violence, data collection method, participant characteristics, inclusion status (full inclusion or route-only), and notes. Data collection was divided equally between the two reviewers, and all data collection was checked by both reviewers.

Quality assessment

For each study included on a full-inclusion basis, we carried out a risk assessment of bias using the Critical Appraisal Skills Programme (CASP) Checklist for Qualitative Research, which provides an assessment of a study’s validity, reliability, and applicability [46]. The following ten domains of the studies were assessed for quality: 1) Was there a clear statement of the aims of the research? 2) Is a qualitative methodology appropriate? 3) Was the research design appropriate to address the aims of the research? 4) Was the recruitment strategy appropriate to the aims of the research? 5) Was the data collected in a way that addressed the research issue? 6) Has the relationship between researcher and participants been considered? 7) Have ethical issues been taken into consideration? 8) Was the data analysis sufficiently rigorous? 9) Is there a clear statement of findings? 10) Is the research valuable? Quality assessment was conducted independently by two reviewers, and any disagreements were recorded, discussed, and resolved. Across studies, the main quality concerns related to limited reporting of researcher-participant reflexivity and ethical considerations, and in some cases, limited descriptions of analytical approaches. The full assessment for each of study included at full-text stage is reported in S4 Table.

Characteristics of Included studies

This systematic review included 27 studies that described or analysed detailed qualitative evidence of sexual violence experienced by migrants during the course of their unauthorised migration journey [5,18,19,33,34,48–68] with full details included in Table 1, and 25 studies that referenced sexual violence although did not include sufficient detail or discussion of the nature of the violence for full inclusion [35,51,69–90] with full details included in Table 2. Nevertheless, this provided insight into the migration routes most commonly studied where sexual violence is a known problem.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of Studies, full inclusion.

https://doi.org/10.1371/journal.pone.0355474.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Characteristics of Studies, route only.

https://doi.org/10.1371/journal.pone.0355474.t002

The reports categorised for full inclusion reported qualitative data from a total of 1,221 migrants and 294 key stakeholders and service providers, plus an additional approximately 100 participants made up of both migrants and stakeholders, where precise counts were not provided.

Across all 52 reports, the most commonly studied routes fell into two broad categories. One category involved the route north from Central America through Mexico towards the USA (n = 18). Most documented instances of violence along this route took place in Mexico during transit. The second category involved routes across Africa towards Europe (n = 12), which featured prominently in the fully included detailed qualitative studies. These included routes from both West Africa (Nigeria-Niger) and East Africa (Horn of Africa-Sudan), transiting through Libya towards Europe. Across these African routes towards the EU, violence was described as saturating the journey, with a particular concentration of evidence relating to violence in Libya. Other routes that were studied included a route across the South of Africa (n = 2), or central and sub-Saharan Africa (n = 7), a South American route through Venezuela towards Chile (n = 2), and between Costa Rica and Brazil (n-1) and a North-East African route though Egypt (n = 2), a route that transited through the USA to Canada (in some cases with a long-term stop in the USA before moving northwards) (n = 1), a route through Turkey (n = 1), a route from Myanmar towards Bangladesh (n = 2), and a route through Calais towards the UK (n = 1). Some studies detailed multiple routes, others were not specific about the location of violence on the journey.

Level 1 descriptive themes

Six descriptive themes were generated by the initial close readings of texts. These were as follows: 1. ‘Agency’, referring to the general appearance of references to migrant agency across the included texts whether these were positive or negative. 2. ‘Economics’, which referred to economic hardship and structural disadvantage that was linked to an outcome of sexual violence. 3. ‘Climate of violence’, which referred to all references to sexual violence that were linked to other forms of extreme violence, brutality and death. 4. ‘Perpetrators’, categorising the identity of the perpetrators detailed across the included studies. 5. ‘Normalisation of violence’ was noted in reference places where sexual violence was accepted as inevitable on a migration journey. And finally 6. ‘Gender’, referring to gender-based or gendered dynamics evident in sexual violence.

Level 2 interpretive codes

The second level of coding was developed in a second round of close reading across the descriptive themes, and produced a synthesis of data extracted from the included texts that offered insight into how the descriptive themes translated and travelled across the studies. This allowed us to review patterns across the treatment of descriptive themes, as well as how the studies, routes, or characteristics of the studies differed across these themes. The interpretive codes are summarised in Table 3:

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Interpretive Code Summary.

https://doi.org/10.1371/journal.pone.0355474.t003

Agency

The descriptive theme of agency generated several insights. Across the included studies, two main framings of agency were identified. The first framing portrayed a lack of agency amongst migrants in transit, although this lack of agency was varied in terms of explicit use of violence. For example, as a migrant in one interview states ‘if we have to get things with sex, then we have to have sex’ [33]. This suggests a transactional element to sexual violence, and an acceptance of the condition of sexual violence, even though the violence is not physically forced. At times, migrants engaged in commercial sex work, but this decision was contextualised in destitution and a lack of alternatives during transit [49–51].

The portrayal of sexual violence as transactional was common [5,51,10,91]. All of these examples were coded as sexual violence for this review, because of their contexts, including clear articulation that sexual contact was unwanted, power imbalance such as in the context of immigration control or reliance on smugglers, threat of greater violence and death, and destitution or absolute economic need.

In other instances, the lack of agency was more directly linked to physical violence:

‘We were travelling as a group of men and women and came across a group of soldiers. They made us sit down and threatened to arrest and deport us. They started looking amongst us and picked the younger women, and raped them not so far from where the group was sitting. It was so painful, and there was nothing we could do’ [52] p.5.

Such examples demonstrate the lack of agency or the lack of alternative options available to migrants. Keygnaert et al [19] report that the impossibility to resist violence or escape from violence was emphasized by both victims and witnesses to violence interviewed in the study.

The second of the two framings of agency was its characterization as a theme that highlighted the methods migrants used to reclaim agency in the context of sexual violence. These included things like dressing ‘like a boy, to escape’ being picked out for abuse [18], p. 12, or agreeing to a ‘relationship’ with captors as a survival strategy. The excerpt in Adeyinka detailing this makes explicit that this was a strategy adopted as a result of violence, rather than a choice:

‘They think I want to f*** that wicked man [smuggler and trafficker] that tear my body with his big thing? I had no choice now. It was the only way to survive’ [18], p.12.

There were also differences across studies in terms of the perception of agency, which we coded as ‘responsibilisation.’ For example, male migrants often perceived women to possess an ability to negotiate because they had the capacity to ‘offer’ sex as a ‘ticket to travel’ [5,33,10]. Similarly, the acquisition of contraception before a journey might be positioned as consent, or the knowledge of sexual violence along the route produces a ‘blame the victim’ mentality [53]. Across the included studies, the various ways agency appears positions that agency as existing only within the limits imposed by significant structural constraints, forms of oppression, and interpersonal power dynamics.

Economics

Economic need was often established as a push factor for migration in the studies, indicating that economic need is linked to irregular migration as a motivation or cause: those in the most extreme economic need have limited options [49]. Economic need was also associated with entrapment in sexual violence, for example:

‘I realised that no one will help me raise my siblings … they are my younger ones, who will take care of them, if not me? It was while I was hustling and looking for a way to make enough money that I met a woman who promised me a job in Libya as a stylist.’ [18], p. 6.

Entrapment might be enforced with violence, or denial of freedom. In some cases in the West African context, it was enforced with juju rituals in which women were taken to a shrine to swear an oath to repay an amount of money determined by the trafficker [18].

Economics were linked to transactional sexual violence, particularly in the context of ‘survival sex’ to meet the most basic of needs. For example, on the central American route an excerpt reports:

‘I had sex with someone, even for an order of tacos in Mexico. It happened to me on the way, too. I had to have sex with someone in exchange for a glass of water’ [49], p.3.

In this way the sexual violence that emerged linked to economic need was characterised by two dynamics. One was economic need leading to migration that produces migrants who travel the most dangerous routes, with or without engaging smugglers, but vulnerable to exploitation along the route. The second is economic need that leads to structurally or circumstantially coerced engagement in sex work or transactional sex during transit.

Climate of violence

The studies outlined many examples of extreme violence and brutality. These were often connected with travel through West and North Africa [18,19,48,54], but were also present elsewhere, including the Central American route [5,55]. It was evident across these studies that death is visible along migrant transit routes, via the visibility of bodies, and witnessing killing. In this way, sexual violence happens in a context of being surrounded by death: threats of violence are reinforced by witnessing those who resist smugglers and traffickers being brutalised and killed or left for dead [34,48,65].

There is a climate of violence surrounding irregular migration routes that is tangible throughout the evidence in the included studies. This is temporal where violence pre-migration is the, or one of multiple, motivators for migration [65]. The temporal aspect is also linked to being on the journey: migrant journeys are not linear or constantly in motion [92], and evidence of the phenomenon not of waiting for an opportunity to move on, but more brutally of kidnapping and imprisonment are evident in some of the excerpts. For example, Chynoweth et al. [48] refer to men being held in containers, and Gebreyesus et al. [34] describe ‘torture camps’ where women can end up trapped:

‘Sarah was already a captive there. She was trapped there for a long time because she could not pay her ransom. She was raped and tortured for many months’ [34], p. 735.

The climate of violence is also spatial in the sense that the journeys traverse hostile terrain where travellers are entirely dependent on smugglers or traffickers for their survival [34,48,65]. For example, an excerpt describes:

‘[t]he traffickers started to rape us on our way to Sinai. Every time we stopped for the night, the truck drivers came to where we, the women, slept, they used torch lights to select the beautiful one amongst us. They took her to a nearby place, raped her, then returned her to our sleeping area’ [56], p. 584.

The hostile and isolated desert route meant that there was no available escape route without risking death.

Perpetrators

While victims and survivors of sexual violence were not exclusively female, the perpetrators of sexual violence across the included studies were exclusively male. The most common perpetrators of sexual violence were the smugglers or traffickers that were transporting migrants, and border guards, police, and other officials [10], who were using their authority to coerce sexual contact. There were varying levels of systematization and various mechanisms of systematization apparent in the accounts. For example, an excerpt from Vogt [57] documents sexual assault at a militarised migration checkpoint:

‘Everyone was there, migration, the army, the Federal Police were there, there were so many of them. And they checked me, and as I was the woman who had come on the train, how they touched and groped me until I urinated on myself. I urinated because I was scared that they were going to rape me…’ [57] p. 212.

The presence of state officials including the police, collaborating in sexual assault, is suggestive that this is not rare: if the behaviour was unusual one would expect it to be hidden. Other examples of state-affiliated perpetrators include rape being used as a condition to pass through an immigration checkpoint [33,58], rape by police or the military while held in detention [18,54], or when in border areas [48,59]. Evidence also suggested collusion between police and criminal gangs, for example Ghebrezghiabher and Motzafi-Haller [56] include examples of police selling migrant women to traffickers. Rubini [54] reports that ‘individuals who paid for victims’ release from detention centres were the ones that sold them to brothels’ [54], p.7. Smugglers delivering migrants to, and entrapping migrants in prostitution was also evidenced [18].

There were multiple accounts of smugglers and traffickers using systematised sexual violence to control migrants. This included imprisonment, rape, sexual enslavement, forced nudity, sexual assault, forced prostitution, and forced perpetration of sex acts on other migrants [19,48,56].

There was evidence that perpetrators were often in a ‘protective’ role and then abused that position. These cases included police who are charged generally with public protection, but also paid smugglers who were hired to facilitate a migration journey [60,61] and also other migrants with whom female migrants entered a mutually beneficial friendship and were then coerced into sexual contact. For example:

‘He told me, “No, I only care for you as a daughter, and I love you as one, I want to help you because I know all you’ve done to come this far, like leaving your daughter behind, so I want to help your daughter and I will help you.” So, I came all the way up here, and when we got here, he became somebody else. He told me to pay him for all he had done for me, and all he had spent on me, so I told him “You know I don’t have any money,” so he abused me, he grabbed me by force…’ [62] p.1268.

There was some recognition in the data that travelling as a couple afforded benefits for both male and female migrants, although these relationships could also lead to different types of vulnerability. Perpetrators were also opportunists, including migrant men [34,49,57,62,63], and people along the route such as truck drivers, taxi drivers, and local citizens who lived along the common travel routes [19,50,52,56,58,64,69].

Normalisation

The included reports provided evidence that sexual violence is known along the route to the extent that it is accepted as inevitable by travellers, understood to be likely and expected on the journey, cross cutting with the theme of agency, where women took measures before and during travel to prevent unwanted pregnancies. As evidenced in Letona et al [49]:

‘[The trip] scared me a lot. They gave me injections: what were [the injections] preventing? …At first, I didn’t ask myself why; why if they [parents] really knew that [rape] could happen to me, why were they sending their daughter there?” [49]p. 3.

As detailed by a migrant in another study ‘In general, the men smuggling us will be forcing girls to have sex at some point along the journey’ [34], p. 730. Infante et al [33] remark that ‘this perception of the inevitability of sexual violence and of the inability to control or prevent it, was prevalent in the discourse of study participants’ [33] p.1153. Sexual violence in many studies was discussed as part of the risk of undertaking a migration journey, expected, and while measures might be taken to mitigate the effects, could not be entirely avoided.

Gender

Research on sexual violence implies an expectation of gendered victimisation, given the long history of violence against women, and an expectation that perpetrators of sexual violence are male. The cultural normalisation of sexual violence against female migrants raises themes of victim-blaming and responsibilisation [33,55]. Yet, perhaps the most significant finding in this regard intersects with the normalisation of sexual violence, which was (unsurprisingly) gendered female. Three reports focused on sexual violence against men [19,48,58], and it was clear in the data that such violence is widespread but not normalised. This is linked to the shame attached to disclosure of violence [48], but reluctant disclosure also leads to ongoing negative physical, mental, and social effects [19,48].

Level 3 interpretive codes

The third level codes were applied only to three of the six descriptive codes: Agency, Climate of Violence, and Perpetrators. These were the codes with the greatest discrepancies in concept translation and framing across the studies, but these discrepancies formed some clear patterns indicating the need to break them down further. The ‘Agency’ code specified two dominant framings of agency across the included studies as detailed above. One tended towards a negative framing, where agency is deconstructed as a form of responsibilisation, or in more political terms, victim-blaming. In these cases, sexual violence was framed as ‘condoned’, with less attention paid to the structural constraints. The second framing of agency was positive where preventative or strategic action adopted by women in the context of (the threat of) sexual violence was framed as resistance. The third level coding recognised this distinction as a pattern whereby the framing of female agency was either positive [49,91] or negative [51,52] as it travelled across studies.

Climate of Violence also produced two sub-categories in the third level coding stage: one was to portray sexual violence as an additional tool used in a brutal toolkit of extreme violence. The other focused on the protective aspects of sexual violence, where the possibility of sexual violence was seen to reduce the likelihood of death. Again, the distinction was portrayed as a pattern across studies, where some studies highlighted the extent of violent brutality and sexual violence was integrated into this narrative [18,19,34]. Other studies focused on the perception of sexual violence as a constrained choice that offered protection, but was accepted because of its embeddedness in a violent migration journey [33].

Finally, the theme of Perpetrators was divided into two sub-themes: Systematic violence referred to studies that acknowledge where dimensions of the system in which migration happens are permissive of sexual violence. In this case, the dynamics of the immigration system itself were constructed as a cause of violence [51,67,65]. Studies recognised the violence of individual perpetrators, but several also focused on immigration systems as a structural cause that highlights the systemic enablement of sexual violence [18,55,57,60,69]. The second sub-theme looked specifically at the perpetrator types at the individual level: for example types included formal (state, police, immigration officers, border control) and semi-formal such as smugglers and traffickers whose power is enabled solely by the immigration system [18,19,33,52,54,55,62] and then informal perpetrators, such as opportunists during the journey who may or may not be migrants themselves [49,57,62,63,66]. This framing focused on individual perpetrators and the relational power dynamics in specified situations of sexual violence. The full hierarchical coding schema can be found in Fig 2.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Hierarchical Coding Schema.

https://doi.org/10.1371/journal.pone.0355474.g002

Agency: Transactional violence and responsibilisation

The first, and most dominant, framing of agency was to foreground the lack of agency on the part of migrant victims, and victim-survivors, of sexual violence. This intersected with other descriptive themes across the studies. For example, where economic need was most prevalent the routes travelled incurred the most danger of violence, including sexual violence. Additionally, the climate of violence and the extreme brutality documented across these studies further emphasised the lack of agency on the part of migrant victims and victim-survivors. Studies included examples of torture, of forced witnessing of sexual violence, of witnessing death that was a consequence of sexual violence, and of witnessing killing where violence was resisted.

However, a framing of agency asserted by migrants, whereby female migrants adopted birth control before travel to avoid unwanted pregnancy, and where migrants entered into partnerships with other migrants for mutual protection, was also present. In some of these cases sexual violence was constructed as transactional. Yet, the spectrum of transactional sexual violence included examples of sex acts to avoid destitution for needs as basic as food and water. It also included transactions to permit border crossing, and it included engagement in sex work. The acceptance of sexual violence to facilitate border crossing can be understood as constrained agency. The choices for migrants in this setting are limited even when there is no threat of killing or brutality. Condoning or accepting sex acts can be understood as a means of retaining a sense of power: where resistance might result in physical violence, acceptance can retain a sense of agency and dignity. As one woman explained: ‘The advantage of being a woman is that men will help you just to have sex with you. Just think of it as paying for protection with [your] body’ [5].

Geographically, the examples of transactional sexual violence tended towards the Central America – Mexico route, and in these cases men at times perceived women as having an advantage because they could use their bodies to pass through immigration checkpoints, or to appease border guards and police who might otherwise detain and deport them. In these examples, the migratory journey is the objective, and there is a general acceptance of the conditions of travel that are violent and at times brutal. In such cases the descriptive theme of agency intersects with the normalisation of violence on the journey: where the type of unwanted transactional sexual violence would be perceived as shocking in everyday life, on the migration journey it is construed as a means to an end [52]. This evidence is suggestive of the overall violence of the journey: sexual violence can be constructed as a lesser harm to suffer, with the risk of physical harm and death repeatedly demonstrated along the journey. The structural constraints on female migration are enhanced in this way. Women embark on migration journeys for multiple and complex reasons. The agency exercised in the act of irregular migration is subject to the constraint of sexual violence being likely while in transit. This can produce a victim-blaming logic for the lack of protection from sexual violence [10]. For example, women described as a ‘temptation’ for men, removing the agency behind violence from the perpetrator and shifting it to the victim [62]. For other women, sexual violence is posited as an advantage that lessens other physical risks along the route [33]. If sexual violence is positioned as a ‘cost’ of the journey in this way, attention is diverted from both the systemic conditions that permit it, and from the perpetrators,

Economic entrapment and the means of survival

Economic need saturated the migration journey. In cases where poverty was a dominant push-factor for migration, the routes travelled tended to be more dangerous. Several accounts mentioned kidnapping by Bedouin traffickers who commonly used sexual violence and torture, and kidnapping was considered more likely if travelling without the help of a (paid) smuggler. Yet, paid smugglers only offered marginal protection against this type of kidnapping. For example, in one account a migrant described being instructed by a smuggler to bury herself in the sand to avoid kidnapping [34]. At times paid smugglers also sell migrants on to traffickers who then imprison and torture them [34,59]. While money can be a protective factor to ease migration, purchasing better guides, safer routes, and bribes for border crossing, money is not always effective protection.

In some accounts, economic entrapment occurred as a result of money ‘loaned’ for migration. It was relatively common on the West Africa – Libya route for passage to be arranged with the promise of expenses recouped, and the promise was ascertained via ‘juju rituals’ which targeted the family of the migrants as victims and used this threat to ensure compliance in long-term sexual violence during the journey and after reaching the destination country [18,67,93], although the latter was out of the scope of this review. In these cases the economic entrapment is constructed as a direct consequence of the necessary means of travel in order to cross the desert in Niger, and to cross borders without state authorization or the requisite travel documents. The debt incurred to arrange transit is illegal extortion, and entraps women in sexual violence in the form of forced sex work, for which all income services the debt.

Economic hardship was also common along the migration routes. Often, whether travelling with or without a smuggler, journeys can become longer than planned or expected, and money might run out. As detailed above, women report exchanging sex for basic survival needs such as food, water, and shelter [49]. There are also examples of sex being demanded as payment for protection and resources provided by fellow travellers or informal smugglers during a journey and after protection has been delivered [62].

In short, violence is a dimension of poverty [94], and poverty constitutes a risk factor for violence [95]. International irregular migration journeys are no exception. Where poverty is a significant factor in international border crossing, and where this border crossing happens through routes that are least costly, the risk of violence is greater. Yet, violence cannot be explained by poverty in these cases as the risk of violence is high along international migration routes when travelling without state authorization, both because there is little recourse for protection from violence, and because state actors are complicit in violence including sexual violence. Findings suggest that the unauthorised nature of the journeys and the consequent lack of recourse for protection from violence or accountability for perpetrating violence can explain violence on the journey, although poverty is an intersecting factor that increases risk of violence overall.

Climate of violence and the visibility of death

The climate of violence along the route intersects with evidence that violence becomes normalised. However, the evidence for the interpretive theme focused most closely on types of spectacular violence, such as witnessing violent punishment including killing for refusing to comply with sexual violence, and forced witnessing of brutal sexual violence that resulted in death [19,34]. The proximity to violent death on migration journeys is described by participants across the studies.

Proximity to death is spatial in the sense that journeys transit through hostile terrain, often on foot. Travellers who cannot withstand the demands of the journey are often abandoned and left to die by smugglers and the rest of the group. Migrant interview participants describe viewing bodies as they travel through the desert, and the jungle, and seeing people fall from ‘La Bestia,’ the freight train through Mexico on which migrants typically ride atop the carriages [55,57,62]. The proximity to death is a reminder of the climate of violence in which the journey happens, suspending the normal ontological security of everyday life. The reality of hostile terrain reinforces the extent to which migrants rely on the smugglers who support their transit, hence a threat is always implicit in demands from smugglers: failure to comply creates physical risk not just from violence but from the danger of the environment.

Formal and informal perpetrators and contexts of structural violence

The context of structural violence on the irregular migration journey intersects with economic need, a lack of agency, and the climate of violence overall that constructs a proximity to death. In terms of perpetrators, they can be formal and organised such as state actors, semi-formal and organised such as traffickers, and smugglers that are attached to gangs and organised crime. They can also be semi-formal without being organised, where a relationship of need creates a power disparity between a paid smuggler, or a guide, and this power disparity is exploited for sexual violence. Due to the extreme vulnerability constructed during the journey, opportune sexual violence also regularly occurs [62,66,72].

For migrants making irregular journeys, it makes little difference whether perpetrators are representatives of the state such as police or immigration officers, or whether they are organisations such as smugglers, traffickers, or organised criminal gangs. In both contexts the protective role is present but ambivalent (see above findings relating to ‘perpetrators’), and due to the nature of migration without state authorisation, there is a lack of pathway to accountability for violence for the perpetrators, or for the potential for protection from violence for the victims. While there are migrant shelters and safehouses along some of the routes travelled, these can only offer temporary protection. Routes for accountability are limited by lack of legal status in the transit country. Sexual violence often happens during the trajectory of the journey and the migrant victim-survivor continues to travel with the perpetrators. Alternatively it happens while navigating an immigration check and onward travel is gained via submission to sexual violence [52,57]. It is clear that at times there is collusion between state actors and smugglers and traffickers who are responsible for sexual violence [54,56,57]. In these contexts, the irregular nature of the route itself prevents the ability to report violence, to support future prevention of violence, or to seek accountability for violence on the part of the migrant.

There is additionally evidence of systematization of violence in the type and context of abuse. This includes group or collective sexual violence and forced witnessing. Violence is systematised in repeated patterns. There was evidence both that sexual violence can be randomised with groups forced to witness repeated violence creating a situation of terror, and sexual violence can also be associated with punishment for transgression or resistance [87]. Technology-enabled violence was also used in a systematised way to enforce long-term compliance among victims, recording torture and sexual violence to post on social media or threaten with public shame [51]. In these cases, sexual violence is employed as a tool of control and for financial gain. The presence of these patterns in the evidence suggests entrenched and embedded sexual violence in an immigration system that is passively permissive of it, and even benefitting from it in support of efforts to deter undocumented migration.

Normalisation of violence: Gender dimensions

It was well-evidenced, and cut across the various interpretive codes, that violence against women was normalised and expected on the journey. There was very limited data on non-binary border crossers [68], and few studies dedicated specifically to sexual violence against men [58,78,87]. While sexual violence on the journey potentially affects all migrants, the gendered power dynamics shape these experiences [10]. Smugglers, traffickers, and border guards are generally male, meaning women are caught in a power dynamic whereby they rely on men to facilitate border crossing [56]. Sexual violence was discussed in terms that highlighted the inevitability and acceptance of the risk of sexual violence [5,18,10,72]. The most brutal forms of sexual violence, particularly torture and imprisonment over time, were an exception to the normalised language. While women expected to be at risk of sexual violence during the journey, they set out expecting to complete the journey. While this was the case for interviewee victim-survivors, there was a lot of evidence of torture, violence and deaths of women who either did not complete their journey, or there was no reason to believe they completed the journey. This type of sexual violence was widespread on the North African routes [19,34,56], but also discussed in relation to criminal gangs in Central America and Mexico [62].

Where violence is expected it is reported in a way that suggests it is experientially less severe than when it is unexpected. Of course, this is a value judgment guided by perception and may differ significantly by experience. However, considering how the normalisation of violence intersects with agency and the types of (constrained) empowerment adopted by women offers some insight. The most significant finding in this area was the intersection between normalised violence and gender-based violence. Violence against women was widespread, common, and often expected. Sexual violence against men was not as widespread as sexual violence against women, but was still common. This was particularly clear on the North African routes, and evidenced in Libya, but mentioned elsewhere [58,87]. Studies that reported sexual violence against men were initiated via health data, finding physical evidence of sexual violence that launched further qualitative investigation. Violence against women on migration journeys tends to be construed as an expected risk of irregular migration, violence against men remains hidden, associated with shame, and when discussed highlights the extreme brutality. While this reflects gendered power dynamics and female victimhood in general, it also suggests a tactic of violence that emasculates male migrants travelling without authorization in order to control, extort, and victimise them.