We conducted a thematic synthesis [31] of qualitative studies that reported experiences of violence by people in insecure migration status. This report follows the Cochrane guidance for undertaking a systematic review [32] and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) reporting checklist [33]. The protocol was prospectively registered on PROSPERO [CRD42021268772] [34]See Appendix 1.

Eligibility criteria

We included primary studies of any design that used qualitative methods for data collection (e.g., interview, focus group, observation, document review) and analysis (e.g., content, narrative, discourse, thematic, grounded theory), if they documented first person excerpts that described an association between insecure migration status and the experience of violence. Only peer reviewed reports in English published since 1 January 2000 were included.

This review followed a Population-Exposure-Outcome (PEO) design, in which the exposure was insecure status, and the outcome was violence. All participants were migrants who were in a status that embedded a form of insecurity. All participants experienced violence in the context of insecure migration status. To meet the inclusion criteria, the violence had to be linked to the insecure migration status. The analysis traced the link that participants made between the violence they experienced and their insecure migration status. Insecure status was defined according to Innes 2023 [8] and was formed of a spectrum of different statuses (see Appendix 2). These included undocumented, asylum seeking, family-based and employment-based statuses.

The definition of physical violence that was adopted in this study included interpersonal and state physical violence and physically forced sexual violence (rape, sexual assault) as specified in the World Health Organization definition and typology of violence [35]. We included all forms of physical and physically enforced sexual interpersonal violence. We also included state violence, where physical violence and/or physically forced sexual violence was perpetrated by an agent of the state acting in their professional capacity (including border enforcement, police, and immigration officers). The focus on physical and physically enforced sexual violence was not intended to undermine the relevancy of other forms of violence such as structural, systemic, legal, biological, psychological, and emotional. Rather, it was to limit an unwieldy study to the most explicitly violent contexts to offer insight into where physical violence is experienced as linked specifically to insecure migration status.

Search strategy

We combined three concept clusters that were reviewed by the team of researchers, which included expertise in migration studies, violence, and research methods. The concepts clustered terms relating to ‘immigration’, ‘violence’ and ‘methods’. A Boolean search was carried out to link each of the concept clusters with each other (AND search) while using multiple descriptive terms in each of the three clusters (OR search). See Appendix 3 for more details.

Database selection was based on initial scoping, combined with areas of expertise across the authorship. Five databases were selected: Embase, Social Policy and Practice, Political Science Complete, SocINDEX and Web of Science Social Sciences Citation Index. We ran the searches on 22 September 2021 and updated on 31 May 2023, for records from 1 January 2000. The start date was chosen to exclude work that predated immigration reforms in the 1990s. All selected studies were subject to backwards and forwards citation tracking to identify additional studies for inclusion. Forwards citation tracking was carried out using the tool available in Google Scholar.

Study selection and data extraction

Endnote was used to deduplicate the search results and to save PDF files. The first reviewer screened all titles and abstracts against the inclusion and exclusion criteria and studies that satisfied the inclusion criteria at the abstract stage then went forward to full text screening. Full texts were screened against the exclusion matrix (see Appendix 4) and the reason for exclusion was recorded. Both stages of screening were carried out in Rayyan [36]. The second reviewer independently screened 20% at both stages of review and all discrepancies were resolved through discussion and consensus.

Data collection process

Details of each included text were recorded in a bespoke Excel table documenting seven categories: (a) report ID and year, (b) insecure status type, (c) violence type, (d) dataset details, (e) country or region of violence, (f) participant characteristics, and (g) notes. These details were documented by the first reviewer and then checked for accuracy by the second reviewer.

Quality assessment

We carried out a detailed risk of bias assessment of each included study using the Critical Appraisal Skills Programme (CASP) Checklist for Qualitative Research (CASP 2018). We assessed quality for ten domains per study: (1) Was there a clear statement of aims? (2) Is qualitative methodology appropriate? (3) Was the research design appropriate for the aims? (4) Was the recruitment strategy appropriate for the aims? (5) Was the data collected in a way that addressed the research issue? (6) Was the relationship between researcher and participants considered? (7) Were ethical issues considered? (8) Was the data analysis sufficiently rigorous? (9) Was there clear evidence of findings and (10) Is the research valuable? We did not give the study an overall score but reported the complete assessment (see Appendix 5). The quality assessment was carried out independently by two reviewers and any disagreements were discussed, resolved, and recorded.

Synthesis

We adopted a thematic synthesis approach as the most suitable for our research question exploring experiences of violence [31]. All the included reports were imported into NVIVO.

At stage one, first reviewer used a combination of inductive and deductive approaches to code each report line-by-line. The coding strategy was derived through an iterative process after two readings of the included reports. Only first-person descriptions of physical violence that were linked to insecure migration status specifically by the speaker were coded. The link might have been made in contextualized information provided in the article (such as the author stating that they asked the speaker specifically about their insecure migration status). These codes were reviewed by and agreed with the second reviewer who coded 100% of reports. All reports were double coded by the first reviewer to ensure any codes derived through line-by-line coding were assessed for every report, and coded once by the second reviewer. Discrepancies were logged in an Excel table, discussed, and agreed upon.

At stage two, the first reviewer developed analytical themes by reviewing co-occurrence across inductive and deductive codes. The two reviewers discussed the themes before finalizing.

FindingsCharacteristics of the included studies

We included 31 studies, published in 33 reports, reporting qualitative experiential data of a total of 1507 migrant participants (at least 49% female, two studies did not disclose the gender distribution) (Fig. 1 and Table 1). All but two [24, 37] of the studies used a form or a combination of forms of interview methodology. Six studies used ethnographic or participant observation [23, 38,39,40,41,42], three studies used focus groups [18, 40, 43, 44], and three used participatory action research [27, 37, 45].

Table 1 Characteristics of included studies

Ten studies were located in the USA [18, 40,41,42, 46,47,48,49,50], eight in Europe including the UK [23, 24, 26, 27, 37, 38, 51,52,53], two in East Asia [39, 54], four in Africa [45, 55,56,57], four in the Middle East and South Asia [43, 44, 58,59,60], and three in Mexico [61, 62].

Fifteen of the studies linked violence to undocumented status [38, 40,41,42, 45, 46, 49,50,51, 55,56,57,58, 61, 63], seven to spousal sponsorship [18, 26, 27, 47, 48, 52, 54, 64], five to asylum seeking [23, 24, 43, 44, 53, 62] and three to employment-based statuses [39, 59, 60]. One text defined immigration status just as ‘insecure’ [37].

Fig. 1Descriptions and patterns of violence experienced when in insecure migration status

We developed 14 inductive and 4 deductive codes. The inductive codes linked the types of violence that were most commonly experienced in insecure migration status. The deductive codes summarized the perceived link between violence and insecure status: (1) direct insecurity, (2) fear of removal, (3) lack of recourse to state support (e.g. law, refuge, economic), (4) gender (Appendix 6.

We developed four analytical themes summarizing experiences of violence linked to insecure migration status: Vulnerability to Sexual Violence, Lack of Pathway to Support, Power Imbalance and Gender: violence against women (Table 2).

Vulnerability to sexual violence

The analytical theme ‘Vulnerability to Sexual Violence’ was drawn from the data linking physically forced sexual violence to direct experiences of insecurity, that is, the association participants in the included studies made between sexual violence they had experienced that they directly related to their lack of immigration status at the time the violence occurred. In this context, ‘Direct Insecurity’ was the most commonly occurring deductive code, and referred to excerpts that described an exposure to violence that was directly linked to experiences of being in insecure migration status. In this context the overarching insecurity gave rise to violence (rather than the violence being repeated or prolonged as a result of insecurity). This applied primarily to people who were without status and most commonly occurred with the descriptive codes ‘sexual-violence’, ‘in-transit-highly-vulnerable’, ‘in-destination-highly-vulnerable’, and ‘employment-based-violence’. Violence that was associated with the direct insecurity of being without status was primarily evident during undocumented migration journeys, particularly when those journeys were facilitated by smugglers or traffickers. It was also evident after arrival in a destination country if a person remained in undocumented status in the receiving country.

Sexual violence on migration journeys

There were many descriptions of physically enforced sexual violence occurring during migration journeys, which were primarily documented in the dataset through African states on journeys towards Europe, through the Sinai desert towards Israel, and through central America and Mexico towards the US. These descriptions emphasised the lack of power to resist sexual violence on the part of migrant victims. For example, one participant recounts “I had no choice but to give in” [55], and another specified “I felt so powerless with the guns pointed at us” [44]. While extreme violence is described in detail across many of the studies, at times, sexual violence was described as a necessity or as a known burden attached to the migration journey. A level of acceptance of the lack of power to resist is indicated, for example, “it was not easy, and it’s life” [45]. At times the violence associated with the migrant journey was expected and described as transactional. This might be in an immediate sense, for survival as recounted in Adeyinka [55], or as a form of payment for the intended end result of a successful migration: “Even migrants say that women have the ticket to transit between their legs, and the police tell them they have to have sex with them to be able to move on” [63], and “Just think of it as paying for protection with [your] body” [61]. Nevertheless, it is worth noting that the reference to a transaction is not denoting consent.

Sexual violence at destination

Experiences of psychically enforced sexual violence in the destination country were linked to the vulnerability of being without immigration status that saturates all parts of life. In most descriptions of violence in this context, the violence was not expected after arrival at the destination. For example, “I would cry every time they had sex with me, because that was not what they told me that I would come and do here” [55]. Sexual violence in the destination arose from living quarters or from employment arrangements where women were vulnerable because their insecure status meant that they had little power to resist physical violence [43]. They might have been coerced into sex work in order to pay back high fees for transit, with the threat of physical violence or imprisonment if they refuse. For example, one woman recounted being forced to have sex with men who were brought to her room; she was not allowed out in case she ran away [55].

One article detailed ‘marriage trafficking’ cases where women were sold, against their will, into marriage and trapped in the situation by bearing children. These women also described sexual violence, for example:

“That night, I was raped by a cripple. Later, I found out that I was sold to that cripple (Trafficked woman 4). … I was raped by my husband in his house the first night. Then I was locked up by my husband’s family in a room with a big iron door. (Trafficked woman 5)” [54].

These examples are indicative of the body of excerpts that show the types of violence associated with direct insecurity, conceptualised as a lack of immigration status or an explicitly insecure immigration status. Sexual violence occurred most frequently and co-occurred with descriptions of being in transit and with descriptions of insecurity in the destination country. Direct insecurity and sexual violence can also be linked to the cost of transit and the desire for smugglers or traffickers to recoup the costs, such as in the case of marriage trafficking and forced prostitution. This vulnerability to sexual violence in particular was clearly associated by participants across several studies with their explicitly insecure, undocumented status [38, 39, 43,44,45, 51, 54, 55,