Exploring the impact of ‘hostile environment’ policies on psychological distress of ethnic groups in the UK: a differences-in-differences analysis

doi:10.21203/rs.3.rs-3423720/v1

Purpose

In 2012, the UK government announced policies designed to create a ‘hostile environment for illegal migration.’ This included sweeping changes immigration legislation which included the introduction of immigration controls enforced by employers and landlords and anti-migrant rhetoric in the press. In this paper, we measured changes in psychological distress among people from ethnic minoritised groups compared to White British controls from prior to the introduction of the hostile environment policies through their implementation.

Methods

We used Understanding Society, a UK longitudinal household survey from 2009 to 2020 (Waves 1 to 10). We included six ethnic groups: Bangladeshi, Black African, Black Caribbean, Indian, Pakistani, and White British. We used difference-in-difference models to estimate the marginal mean psychological distress score, as measured by the General Health Questionnaire (GHQ-12) during the three eras: pre-policy era (2009–2012); (2) transition era (2012–2016); and (3) ongoing policy era (2016–2020).

Results

In the pre-policy era, we found elevated levels of psychological distress (marginal mean = 2.70), Bangladeshi and Caribbean groups (marginal mean = 2.47) compared to the White British group (2.08). Psychological distress increased during the transition era for the Pakistani and Bangladeshi groups (2.92 and 2.71, respectively), and continued to rise in the ongoing policy era for the Bangladeshi group (2.80). We observed lower psychological distress in the African group in the pre-policy era (2.16), decreasing in the transition era (1.91), and a small increase in the ongoing policy era (1.95). The Indian group showed similar patterns to the African group. In contrast, psychological distress remained stable across all three eras for the White British group.

Conclusion

Psychological distress increased in Pakistani and Bangladeshi individuals following the introduction of hostile environment policies. We did not find an impact for Indian, African, or Caribbean groups. This finding underscores the importance of disaggregating analyses by ethnic group to capture the distinct identities and experiences and how these intersect with mental health.

common mental disorders

depression

anxiety

psychiatric epidemiology

temporal trends

difference-in-differences

ethnic minoritised groups

In 2012, then Home Secretary Theresa May announced a desire ‘to create here in Britain a really hostile environment for illegal migration’, which has later been referred to by the Government as the ‘compliant environment policy’. The passage of the Immigration Acts of 2014 and 2016 enacted this ‘hostile environment’ by implementing several policy changes that aimed to increase social exclusion for undocumented migrants. Alongside these policy changes, other initiatives began such as the ‘Go Home vans’, a 2013 government campaign in which vans with advertising slogans reading ‘In the UK illegally? Go home or face arrest’ drove around six London boroughs with high concentrations of immigrants and minoritised ethnic individuals. Further, anti-migrant rhetoric increased in the lead up to the 2016 Brexit referendum, and a spike in racist or religious abuse hate crimes reported to police was observed in the period directly following the referendum, with evidence supporting a causal link with the referendum (1).

The enforcement of national borders was extended beyond traditional means to dissuade illegal residence in the UK and prevent migrants from accessing basic services, employment opportunities, or housing. Griffiths and Yeo describe this process as the ‘deputisation’ of immigration control, wherein the responsibility of enacting immigration policy is diffused to a range of actors, including public servants, medical practitioners private companies, and ‘ordinary people’(2). Within the health system, identification checks were introduced for patients accessing non-emergency hospital care, and a memorandum of understanding between the Department of Health, National Health Service, and the Home Office allowed for the sharing of non-clinical data ‘to support effective immigration enforcement’(3, 4). Further, the Immigration Act 2016 placed immigration control in the hands of employers and landlords, making them criminally responsible for employing or renting to someone that they have reasonable cause to believe is not permitted to work or reside in the UK (2). Such deputisation leads to third parties conducting discriminatory checks and employing restrictions based on ethnicity as they seek to avoid sanctions, negatively impacting both migrants and lawful UK residents from minority ethnic backgrounds (5). In this way, the hostile environment enforces structural racism by encouraging racial profiling through discriminatory checks against anyone ‘foreign-looking’ (2). Indeed, ethnic minoritised groups have been disproportionately subject to immigration checks and denied access to basic services; for example, evidence indicates that since the advent of the hostile environment policies, landlords are less likely to rent to people who are foreign nationals or ethnic minoritised individuals without British passports (2, 5–7).

By 2018, the Government acknowledged the failings of the hostile environment policies. While these policies initially targeted undocumented and irregular migration, they impacted members of ethnic minoritised groups, including those with the legal right to reside in the UK.

This widespread structural discrimination faced by UK residents belonging to ethnic minoritised groups in the last ten years poses a risk to their mental health (Qureshi et al., 2020); people belonging to ethnic minority groups in the UK who have reported experiences of perceived racial discrimination are more likely to have greater psychological distress (8–10), with evidence of a cumulative effect (11). Exposure to individual and structural discrimination directly impacts mental health by acting as a chronic psychosocial stressor, activating the biological stress response (12). Structural racism additionally indirectly impacts mental health by unjustly allocating societal resources on the basis of ethnicity, leading to material insecurity and further stress (13).

Ethnic groups are not homogenous, and those belonging to ethnic minoritised groups represent a diverse population with varied experiences of interpersonal, structural, and institutional racism. These differing experiences of racism are likely to result in a unique relationship with mental health problems between ethnic groups. Nevertheless, there is evidence that the mental health of certain ethnic groups in the UK is already worse than White British people (14); for example, common mental disorders are more prevalent among Black women (15). In light of the hostile environment and increases in racial discrimination, these mental health disparities may widen. However, to date, most research on racism and mental health has taken place in the US, and no study has examined the population impact of the hostile environment on the mental health of ethnic minoritised groups in the UK.

Aims

In this study, we aimed to measure changes in psychological distress among people from ethnic minoritised groups compared to White British controls following the introduction of hostile environment policies. We hypothesised that we would see an increase in psychological distress over the study period, particularly in minoritised ethnic groups.

Data source: We used data from Understanding Society, the UK Household Longitudinal Survey (UKHLS), which collects information annually. Approximately 40,000 households were sampled at Wave 1 (2009-10) through a stratified, clustered equal probability sampling process. Postcode sectors were randomly selected from geographical strata as the primary sampling unit (PSU), and households were randomly selected from the PSUs (16). An Ethnic Minority Boost (EMB) sample was incorporated at Wave 1 of UKHLS to provide at least 1,000 adult interviews from important minority ethnic groups. We used data collected from adults (ages 16 or older) who identify as belonging to one of the following six ethnic groups: Bangladeshi, Black African, Black Caribbean, Indian, Pakistani, and White British. Participants were followed over 10 successive data collection waves of UKHLS covering the years 2009–2020 (Wave 1 through Wave 10). We excluded participants from Northern Ireland, as they did not contribute to the EMB. We additionally excluded those who were proxy respondents at Wave 1, as they did not provide mental health data at most study waves.

Exposure: The primary exposure in this study was exposure to hostile environment policies, which were introduced in the UK between 2012 and 2016. We divided the follow-up time into three eras: (1) pre-hostile environment (2009–2012, including Waves 1–3;) (2) transition era (2012–2016, including Waves 4–6); and (3) ongoing hostile environment policies (2016–2020, including Waves 7–9).

Outcome

Self-reported symptoms of psychological distress were measured using the General Health Questionnaire (GHQ-12). This 12-item questionnaire asks about happiness, symptoms of depression and anxiety, and sleep disturbance and is a well-validated tool for estimating psychological distress (17). GHQ-12 scores were calculated at each wave of UKHLS using the standard GHQ-12 scoring method, where the four response options for each item were dichotomised (0 or 1), where higher scores indicated more psychological distress (18).

Covariates

We considered several covariates in our analysis, including age (continuous), sex (female; male), marital status (partnered; separated/divorced/widowed; single), educational qualifications (A-levels or other post-secondary qualifications; GCSC or high school diploma or lower), citizenship status (British citizen; not British citizen), and urban or rural residence. We took the measures of these covariates from each wave of UKHLS to take into account their time-varying nature, with the exception of sex, which we considered to be time-stable.

Analysis:

We described the sample characteristics at baseline (Wave 1) by ethnic group. We used difference-in-difference (DiD) models to estimate the effect of the hostile environment on mental health by comparing the change in mental health before and after the introduction of these policies in ethnic minoritised groups (treated group) compared to the White British group (control). DiD models are widely used to estimate the causal effects of an intervention in longitudinal observational data. This DiD model allowed us to estimate the potential treatment effect (hostile environment policies) on the outcome (mental health) by comparing change before, during, and after the introduction of these policies on the treated group (minoritised ethnic groups) compared to the control group (White British group).

We used mixed models with an interaction term between era and ethnicity to estimate the relationship between ethnic group and mental health during the three eras of policy implementation. Mixed models allowed us to account for our longitudinal data, with the waves of UKHLS clustered within each individual over time. We also accounted for the sampling process of UKHLS, with individuals clustering within primary sampling unit (PSU). The analysis was weighted at the level of the individual to account for sampling probability, differential non-response, and sampling error.

We calculated the marginal mean GHQ-12 score by ethnic group for each of the three policy periods. The marginal mean of GHQ is the predicted mean based on the mixed model if all the observations were treated as if they were fixed at a particular level of ethnicity and policy period; for example, the marginal mean of GHQ for the Bangladeshi group in the pre-policy period is the predicted mean GHQ if all observations in the cohort were treated as if they were Bangladeshi and in the pre-policy period.

The data was assumed to be missing at random, and we used multiple imputation with chained equations to impute missing GHQ and covariate data. Analysis based on imputed datasets was used for our primary analysis and the results from a complete case analysis were included in the online supplement (Supplement A). We used Stata 17.0 (MP-parallel edition) to conduct all analyses.

Descriptive statistics

There were 42,968 participants included in our sample, most of whom were White British (n = 35,918; 84.0%). All other ethnic groups had between 1,132 participants (Bangladeshi) and 1,905 participants (Indian). There were more females than males in our sample, which was most pronounced in among Caribbean (60.1% female), White British (56.4% female), and African (55.9% female) participants. In contrast, there was nearly equal proportions of females and males in the Bangladeshi (49.2% female) and Indian (48.7% female) participants (Table 1).

Table 1 Sample characteristics by ethnic group, 2009-10 (n= 42,968)

	African		Bangladeshi		Caribbean		Indian		Pakistani		White British
	n	%	n	%	n	%	n	%	n	%	n	%
Total
	946	100.0	622	100.0	799	100.0	1,363	100.0	906	100.0	31,407	100.0
Sex
Female	830	57.7	557	49.2	683	60.1	927	48.7	759	52.8	20,253	56.4
Male	608	42.3	575	50.8	454	39.9	978	51.3	679	47.2	15,665	43.6
Age
Mean (SD)	36.2	12.8	33.8	12.8	45.8	16.9	39.5	15.4	35.9	14.2	48.4	18.4
Education
A levels or higher	876	60.9	450	39.8	448	39.4	1,158	60.8	623	43.3	13,756	38.3
GCSE or lower	560	38.9	680	60.1	687	60.4	745	39.1	815	56.7	22,138	61.6
Missing	2	0.1	2	0.2	2	0.2	2	0.1	0	0	24	0.1
British citizen
Yes	858	59.7	909	80.3	1,036	91.1	1,384	72.6	1,171	81.4	35,852	99.8
No	879	40.2	222	19.6	101	8.9	520	27.3	267	18.6	61	0.17
Missing	2	0.1	1	0.1	0	0	1	0.05	0	0	5	0.01
Urban/rural
Urban	1,430	99.4	1,126	99.5	1,131	99.5	1,876	98.5	1,433	99.6	26,648	74.2
Rural	8	0.6	6	0.5	6	0.5	29	1.5	5	0.4	9,270	25.8
Marital status
Partnered (married / cohabiting)	664	46.2	716	63.2	420	36.9	1,265	66.4	946	65.8	22,787	63.4
Separated / divorced / widowed	199	13.8	84	4.7	222	19.5	138	7.2	143	9.9	5,877	16.4
Single	575	40.0	331	29.2	493	43.4	502	26.4	348	24.2	7,244	20.2
Missing	-	-	1	0.1	2	0.2	0	0	1	0.1	10	0.0
Psychological distress
GHQ-12 mean score (SD)	1.86	2.81	2.24	2.94	2.01	2.95	1.78	2.85	2.22	3.19	1.73	2.87

n: number; SD: standard deviation; GHQ: General Health Questionnaire

[Table 1 here]

There were interesting patterns of education across the ethnic groups. In the African and Indian groups, over 60% had completed A levels or higher, with 38.9% (African) and 39.1% (Indian) leaving school with GCSE qualifications or lower. These proportions were the inverse of what was seen in the White British and Bangladeshi groups, where only 38.3% (White British) and 39.8% (Bangladeshi) had achieved post-secondary qualifications (A levels or higher).

The proportions who had British citizenship also varied greatly across ethnic groups. 99.8% of the White British group reported being a citizen of the UK. Several other ethnic groups had high proportions of citizens, including 91.1% of Caribbean participants, 81.4% of Pakistani participants, and 80.3% of Bangladeshi participants. The lowest proportion of citizenship was observed in African participants, with 59.7% reported having citizenship in the UK.

Over 99% of participants from African, Bangladeshi, Caribbean, Indian, and Pakistani groups lived in urban areas. There was a smaller proportion of the White British living in urban areas (74.2%).

Over 60% of participants in Bangladeshi, Indian, Pakistani, and White British groups were married or cohabitating. There were lower rates of marriage/cohabitation in the African (46.2%) or Caribbean (36.9%) groups.

At the start of the study, levels of psychological distress were highest in the Bangladeshi (mean GHQ: 2.24, SD: 2.94) and Pakistani (mean GHQ: 2.22, SD: 3.19) ethnic groups. The lowest levels of psychological distress were observed in the White British ethnic group (mean GHQ: 1.72, SD: 2.87) and Indian ethnic group (mean GHQ:1.78, SD: 2.85).

Difference-in-differences models: We used difference-in-difference (DiD) models to estimate the marginal mean psychological distress score for each ethnic group during the three policy eras: pre-policy era, transition era, and ongoing policy era (Table 2). Figure 1 displays the trends in psychological distress based on the marginal means from the adjusted model across the three policy eras by ethnic group. During the pre-policy era, the highest marginal mean GHQ was found in the Pakistani group, which had a marginal mean GHQ of 2.70, compared to 2.08 in the White British group. The Bangladeshi and Caribbean groups similarly had elevated GHQ scores of 2.47. The marginal mean GHQ of both the Indian and African ethnic groups in the pre-policy era were similar to the White British group, at 2.13 and 2.16, respectively.

Table 2

Psychological distress scores by ethnic group and policy period, imputed data
	PRE-POLICY PERIOD			TRANSITION PERIOD			ONGOING POLICY PERIOD
	Coeff.	95% CI		Coeff.	95% CI		Coeff.	95% CI
African
Unadjusted	1.94	1.57	2.30	1.59	0.93	2.24	1.78	1.06	2.50
Adjusted¹	2.16	1.77	2.54	1.91	1.24	2.58	1.95	1.23	2.68
Bangladeshi
Unadjusted	2.23	1.76	2.70	2.46	1.58	3.34	2.57	1.61	3.53
Adjusted¹	2.47	2.00	2.94	2.71	1.85	3.57	2.80	1.79	3.82
Caribbean
Unadjusted	2.25	1.84	2.67	2.18	1.46	2.90	2.13	1.38	2.88
Adjusted¹	2.47	2.03	2.91	2.46	1.72	3.20	2.36	1.57	3.16
Indian
Unadjusted	1.84	1.50	2.17	1.65	1.10	2.19	1.63	1.04	2.22
Adjusted¹	2.13	1.78	2.48	1.98	1.41	2.55	2.02	1.42	2.62
Pakistani
Unadjusted	2.55	2.11	2.98	2.64	1.85	3.42	2.42	1.58	3.25
Adjusted¹	2.70	2.26	3.14	2.92	2.16	3.68	2.80	1.96	3.64
White British
Unadjusted	1.85	1.78	1.91	1.85	1.74	1.95	1.87	1.76	1.98
Adjusted¹	2.08	1.98	2.19	2.09	1.95	2.24	2.09	1.93	2.24
¹Adjusted for sex, age, marital status, education, citizenship status, urban/rural status.

[Table 2 here]

During the transition era, the marginal mean GHQ score increased for the Pakistani and Bangladeshi groups, reaching 2.92 and 2.71, respectively. The remaining ethnic groups saw their GHQ scores remain stable or decrease slightly, with the greatest decrease seen in the African group to a marginal mean GHQ score of 1.91.

Finally, in the ongoing policy era, the marginal mean GHQ score continued to rise for the Bangladeshi group, reaching 2.80. The Indian and African groups saw slight increases in their marginal mean GHQ scores, reaching 2.02 and 1.95, respectively. The marginal mean GHQ score in the Pakistani and Caribbean groups decreased to 2.80 and 2.36, respectively, while the GHQ score of the White British group remained approximately stable at 2.09. Similar patterns were found in the complete case analysis (Supplement).

[Figure 1a-e here]

In this first quantitative assessment of the potential impact of the hostile environment policies on the mental health of different ethnic groups in the UK; we found that the hostile environment was associated with negative impact on the mental health of Pakistani and Bangladeshi groups. These ethnic minoritised groups experienced greater psychological distress during and after the implementation of the hostile environment policies, compared to the pre-policy period and compared to the White British group. We found no evidence of negative impact on mental health for the Indian, African, and Caribbean groups. There was no overall common effect of the hostile environment on psychological distress across the six ethnic groups included in our study. This finding underlines the importance of considering ethnic groups not as a monolith, but instead as groups with distinct identities and relationships with mental health.

The differences in the mental health impact of the hostile environment between the six ethnic groups may be partially explained by intersections of ethnicity with other socioeconomic factors. We found the highest rates of psychological distress among Pakistani and Bangladeshi people. Pakistani and Bangladeshi people have the lowest household incomes, the lowest employment rates, and the lowest median gross hourly pay among ethnic groups in the UK (19); long-term trends show that this income inequality for Pakistani and Bangladeshi people has persisted over the last 15 years (20). Reflecting our findings, Bamford et al. (2021) similarly found the highest odds of psychological distress among Pakistani people in a study on ethnicity and mental health in the UK during the hostile environment transition era (21). They examined different risk factors and found that gender, economic insecurity, and the length of settlement in the UK may play a role in the mental health of different ethnic groups (21). Similarly, Wallace et al. (2016) found that the poor mental health of Pakistani and Bangladeshi people in the UK during the same time period was partially explained by racial discrimination and socioeconomic disadvantage (11). The pre-existing economic insecurity faced by Pakistani and Bangladeshi people in the UK may mean they have limited resources to weather potential increased insecurity due to hostile environment policies.

Identifying a causal effect of the hostile environment within ethnic groups is complicated by the measurement of the exposure accurately and cumulatively over time. The hostile environment is in fact a collection of several different policies, which each might have different and perhaps opposing relationships on the different ethnic groups and subgroups included in this study. For example, data from 2016 to 2018 shows that Indian households were the most likely to be owner occupiers out of all ethnic groups in England, suggesting that a relatively low proportion of Indian people have been subject to discrimination in the private rental sector (22). Pakistani, Bangladeshi, Black African and Black Caribbean people are more likely to experience insecure work, indicating that they may face greater job insecurity in light of discriminatory employment restrictions (23). The heterogeneity of hostile environment experiences underscores the importance of investigating ethnic minoritised groups separately.

The GHQ-12 is a multidimensional measure of mental health, which includes anxiety and depression, which are relevant to assessing the impact of hostile environment policies. However, there are some limitations with using this measure, particularly in sensitivity to measure change over long periods of time. In our study, the White British group showed no meaningful change in GHQ scores over the decade of follow-up, suggesting the measure may not be sensitive to change over extended time periods. However, self-reported measures of mental health, including the GHQ, do present advantages for measuring mental health over alternative methods, such as health services records, which can underestimate mental health problems. Some ethnic groups are less likely to access mental health services due to cultural beliefs, stigmatisation, language barriers, and discrimination from the health care system (24), reducing their rate of diagnosis. Nevertheless, future research may consider alternative outcome measures. Effect sizes for GHQ also appear to be relatively small. However, a score of 2/3 has been found to be the most appropriate cut-off and is consistent with the diagnosis of a common mental disorder (17); almost all of the ethnic groups in our study had a mean GHQ score over 2 across the three eras, and the Pakistani and Bangladeshi groups had GHQ scores close to 3 during the transition and ongoing policy eras. Furthermore, while an increase of a few decimal points in a GHQ score may be inconsequential at the individual level, an increase of 0.33 points in the marginal mean GHQ (as seen in the Bangladeshi group) is equivalent to shifting the entire population of an ethnic group to a higher level of psychological distress, as described by Geoffrey Rose (25). Even small changes in GHQ at the population level could indicate a meaningful increase in mental health problems experienced in that population group.

Our study was limited by relatively small sample sizes for the ethnic minoritised groups, reflected in the wide confidence intervals for our estimates. To take into account further heterogeneity within ethnic groups, such as stratification by gender, our sample size would be reduced, further limiting our ability to detect an effect. Future research needs to be conducted using surveys that robustly sample ethnic groups, as there are clear limitations to a sub-group analysis using a general population survey. The small sample sizes may explain why no meaningful effect of the hostile environment was detected for the Black Caribbean group. Considering the particular discrimination and risk of deportation faced by this ethnic group (26), this finding was surprising, and requires further investigation. We were also not able to disaggregate findings by migrant status, and it is plausible that there may be distinct patterns of psychological distress based on migrant generation (e.g. immigrants and second-generation migrants).

Finally, we conducted a mixed effects model to account for the complex sampling structure of UKLHS. However, the introduction of a random effect for PSU means that our model cannot eliminate the potential bias of unmeasured confounding of time-stable variables between individuals (27). Therefore, while we adjusted our DiD model using several time-varying covariates, our estimates are not necessarily reliable causal estimates for the effect of the hostile environment on ethnic groups if our assumption of no unmeasured confounding is violated, although the potential bias is reduced in the presence of more time points (27). However, there is also a risk of over-adjustment in such an analysis, as the hostile environment policies may have acted through many variables to impact mental health. We therefore chose not to control for particular time-varying socioeconomic variables, notably income and employment, because they were assumed to be on the causal pathway as mediators in the relationship between the hostile environment policies and mental health; some of the hostile environment policies had a direct impact on employment for ethnic minoritised groups by increasing discriminatory “right to employment” checks, so adjusting for employment would bias our estimates of the mental health impact.

Overall, our findings underline the importance of disaggregating ethnic minoritised groups and highlight a potential impact of hostile environment policies on Pakistani and Bangladeshi groups in the UK.

Acknowledgements

This study was funded by the National Institute for Health Research (NIHR) School for Public Health Research (Grant reference: BH154142) and SPHR ResNet grant (JD). The views expressed are those of the author(s) and not necessarily those of the NIHR or the Department of Health and Social Care. JD, MM, KD, and DO are supported by the NIHR University College London Hospital Biomedical Research Centre. DO is supported by NIHR Applied Research Collaborative North Thames. JD receives support from the NIHR Advanced Fellowship (Grant reference: NIHR302266).

Competing interests

The authors have no conflicts of interest to declare.

Clifton-Sprigg J, James J, Vujic S. Love Thy Neighbour? Brexit and Hate Crime. SSRN Electron J [Internet]. 2020 [cited 2023 Jan 26]; Available from: https://www.ssrn.com/abstract=3738844
Griffiths M, Yeo C. The UK’s hostile environment: Deputising immigration control. Crit Soc Policy. 2021;41(4):521–44.
Hiam L, Steele S, McKee M. Creating a ‘hostile environment for migrants’: the British government’s use of health service data to restrict immigration is a very bad idea. Health Econ Policy Law. 2018;13(2):107–17.
Potter J. Border control in a healthcare setting is not in the public’s best interests. J Public Health. 2017;39(2):219–20.
Qureshi A, Morris M, Mort L. Access denied: the human impact of the hostile environment. London, UK: Institute for Public Policy Research; 2020 Sep.
Shelter [Internet]. 2016 [cited 2023 Jan 26]. Right to Rent roll out: one more barrier to finding a home? Available from: https://blog.shelter.org.uk/2016/01/right-to-rent-roll-out-one-more-barrier-to-finding-a-home/
Grant S, Peel C. “No Passport Equals No Home”: An independent evaluation of the ‘Right to Rent’ scheme [Internet]. Joint Council for the Welfare of Immigrants; 2015 Sep [cited 2023 Aug 31]. Available from: https://jcwi.org.uk/sites/default/files/documets/No%20Passport%20Equals%20No%20Home%20Right%20to%20Rent %20Independent%20Evaluation_0.pdf
Hackett RA, Ronaldson A, Bhui K, Steptoe A, Jackson SE. Racial discrimination and health: a prospective study of ethnic minorities in the United Kingdom. BMC Public Health. 2020;20(1):1652.
Paradies Y, Ben J, Denson N, Elias A, Priest N, Pieterse A, et al. Racism as a determinant of health: A systematic review and meta-analysis. PLoS ONE. 2015;10(9):e0138511.
Karlsen S, Nazroo JY, McKenzie, K, Bhui K, Weich S. Racism, psychosis and common mental disorder among ethnic minority groups in England. Psychol Med. 2005;35(12):1795–803.
Wallace S, Nazroo J, Bécares L. Cumulative Effect of Racial Discrimination on the Mental Health of Ethnic Minorities in the United Kingdom. Am J Public Health. 2016;106(7):1294–300.
Berger M, Sarnyai Z. ‘More than skin deep’: stress neurobiology and mental health consequences of racial discrimination. Stress Amst Neth. 2015;18(1):1–10.
Williams DR, Mohammed SA. Racism and health I: Pathways and scientific evidence. Am Behav Sci. 2013;57(8):1152–73.
Ahmad G, McManus S, Cooper C, Hatch SL, Das-Munshi J. Prevalence of common mental disorders and treatment receipt for people from ethnic minority backgrounds in England: repeated cross-sectional surveys of the general population in 2007 and 2014. Br J Psychiatry. 2022;221(3):520–7.
McManus S, Bebbington E, Jenkins R, Brugha T. Mental health and wellbeing in England: Adult Psychiatric Morbidity Survey [Internet]. Leeds, UK: NHS Digital; 2014 p. 1–405. Available from: https://assets.publishing.service.gov.uk/government/uploads/system/ uploads/attachment_data/file/556596 /apms-2014-full-rpt.pdf
Institute for Social and Economic Research. Understanding Society: Waves 1–12, 2009-2021and Harmonised BHPS: Waves 1–18, 1991–2009, User Guide. Colchester: University of Essex; 2022 Nov.
Goldberg DP, Gater R, Sartorius N, Ustun TB, Piccinelli M, Gureje O, et al. The validity of two versions of the GHQ in the WHO study of mental illness in general health care. Psychol Med. 1997;27(1):191–7.
Goldberg DP, Williams P. A user’s guide to the General Health Questionnarie. Slough, UK: NFER-Nelson;
Francis-Devine B. Which ethnic groups are most affected by income inequality? 2020 Aug 10 [cited 2023 Aug 31]; Available from: https://commonslibrary.parliament.uk/income-inequality-by-ethnic-group/
Handscomb K, Henehan K, Try L. The Living Standards Audit 202 [Internet]. London, UK: Resolution Foundation; 2021 Jul [cited 2023 Aug 31]. Available from: https://www.resolutionfoundation.org/publications/the-living-standards-audit-2021/#:~:text=In%20this%20report%2C%20our%20latest,mean%20for%20future%20living%20standards.
Bamford J, Klabbers G, Curran E, Rosato M, Leavey G. Social Capital and Mental Health Among Black and Minority Ethnic Groups in the UK. J Immigr Minor Health. 2021;23(3):502–10.
Ministry of Housing, Communities and Local Government. Home ownership [Internet]. 2020 [cited 2023 Aug 31]. Available from: https://www.ethnicity-facts-figures.service.gov.uk/housing/owning-and-renting/home-ownership/latest#by-ethnicity
The Health Foundation. Percentage of employees in insecure work by region/nation, sex, age, ethnicity and income - The Health Foundation [Internet]. 2022 [cited 2023 Aug 31]. Available from: https://www.health.org.uk/evidence-hub/work/job-security/percentage-of-employees-in-involuntary-insecure-work-by-region-sex-age-ethnicity-income
Memon A, Taylor K, Mohebati LM, Sundin J, Cooper M, Scanlon T, et al. Perceived barriers to accessing mental health services among black and minority ethnic (BME) communities: a qualitative study in Southeast England. BMJ Open. 2016;6(11):e012337.
Rose G. Sick Individuals and Sick Populations. Int J Epidemiol.:7.
Slaven M. The Windrush Scandal and the individualization of postcolonial immigration control in Britain. Ethn Racial Stud. 2022;45(16):49–71.
Gunasekara FI, Richardson K, Carter K, Blakely T. Fixed effects analysis of repeated measures data. Int J Epidemiol. 2014;43(1):264–9.

No competing interests reported.

Supplement.docx

Exploring the impact of ‘hostile environment’ policies on psychological distress of ethnic groups in the UK: a differences-in-differences analysis

Status:

Journal Publication

Version 1

Abstract

Purpose

Methods

Results

Conclusion

Figures

Introduction

Methods

Analysis:

Results

Discussion

Declarations

Acknowledgements

References

Additional Declarations

Supplementary Files

Status:

Journal Publication

Version 1