1 Introduction

The objective of this paper is to explore whether individuals born outside the UK were more likely to receive Universal Credit in wave 13 of UKHLS. One of the most important and contentious issues in British public debate is “the impact immigration has on the tax and welfare system” (Frattini & Dustmann,2014). The perception immigrants ‘free ride on benefits’ or are disproprionately “dependent” on out-of-work state support programs, is a prominent factor in the perpetuatation of xenophobic views. (Drinkwater & Robinson, 2011, pp 2-4).Since the 1950s, the growing influx of immigration into the UK has precipitated a raft of restrictions on both immigration and social benefits to weaken, in the words of David Cameron, the “magnetic pull of Britain’s benefit system” (Mason, 2014). Furthermore, the UK’s entry into the single market was followed by great fears that ‘benefit tourism’ had ‘overburdened’ the welfare system. Grievances appropos of immigrants’ access to social protection is even argued to be have been a “key factor” in the 2016 Brexit referendum result(Bertolini & Clegg, 2020, p 419-432). Hence, analysis on whether immigrants are more likely to receive benefits than native born Britons is vital for public debate and policymaking. However, the literature is ‘mixed’, or contested, on whether immigrants disproportionately claim social benefits, and particularly unemployment benefits. A 2017 House of Commons Briefing Paper claims people born outside the UK are less likely to receive “key DWP out-of-work benefits than UK nationals”. It sites a Labour Force Survey which found 7.4% of people born outside the UK in February 2016 received these benefits: despite constituting 17.6% of the working age population (Keen & Apostolova, 2017, p 5).(Frattini & Dustmann,2014) found immigrants since 2000 were 43% less likely to receive state benefits or tax credits and made net contributions to public finances “regardless of area of origin”. However, Migration Watch UK in 2016 claimed in 2014/15 both EEA and non-EEA migrants were a net fiscal cost to the public finances(Vargas-Silva, Sumption, & Walsh, 2022).

Some of these discrepancies are explained by differences in how immigrants are defined, categorised, and measured. This paper operates under the definition employed by the census, Labour Force Survey and Annual Population Survey in the UK. An immigrant is “foreign born”: born outside the country under study(Anderson & Blinder, 2024). Furthermore, this paper focuses on Universal Credit (UC). Introduced in the 2012 Welfare Reform Act, UC is the first “general scheme of guaranteed minimum resources” for people of working age in the UK since the abolition of the Supplementary Benefit. UC is a household entitlement, paid to people out of work or with a low income, which is gradually replacing six-means tested “legacy benefits” (Bertolini & Clegg, 2020, p 429). In December 2023 there were 6.3m people claiming UC(National-Audit-Office, 2024). A comprehensive investigation into whether immigrants disproportionately receive UC is a significant gap in the literature. Assessing whether UC is a successful innovation, and is returning ‘value for money’, at least in the public debate, will partly be determined by whether UC disproportionately is claimed by Britain’s immigrant population. The data used for this analysis is from Understanding Society: the UK Household Longitudinal Study (Institute-for-Social-and-Economic-Research, 2023). Moreover, only wave 13 of data collection, which took place between January 2021 and May 2023, is analysed. 27549 respondents mentioned or did not mention receiving universal credit (with 1483 receiving UC) in wave 13 of UKHLS. Hence, the rigorous data collection methods, large sample size and range of variables on participant qualities makes UKLHS the perfect data source for analysing whether immigrants in the UK disproportionately receive UC.

2 Data and Recoding Variables

m_indresp <-  vroom("/Users/eliotflynn/Downloads/R Files/Data Analysis 3/UKDA-6614-tab/tab/ukhls/m_indresp.tab", 
                   col_select = c("m_benbase4", "m_age_dv", "pidp"))
xwavedat <- vroom("/Users/eliotflynn/Downloads/R Files/Data Analysis 3/UKDA-6614-tab/tab/ukhls/xwavedat.tab", 
                  col_select = c("bornuk_dv", "plbornc", "generation", "sex_dv", "pidp"))

data <- left_join(m_indresp, xwavedat, by = "pidp")

data <- data %>%
  mutate(sex_recoded = case_when(
    sex_dv == 1 ~ "male", 
    sex_dv == 2 ~ "female",  
    sex_dv == 0 | sex_dv == -9 ~ NA_character_  
  ))

data <- data |> 
  mutate(UCrec = case_when(
    m_benbase4 == 1 ~ 1,
    m_benbase4 == 0 ~ 0,
    m_benbase4 %in% c(-1, -2, -7) ~ NA_real_,
    TRUE ~ NA_real_ 
  )) 

data <- data |> 
  mutate(immigrant = case_when(
    bornuk_dv == 2 ~ "Immigrant",
    bornuk_dv == 1 ~ "Native Born",
    bornuk_dv == -9 ~ NA_character_,
    TRUE ~ NA_character_ 
  )) 

country_map <- c(
  "-8" = NA, "-2" = NA,
  "5" = "Republic of Ireland", "6" = "France",
  "7" = "Germany", "10" = "Poland",
  "12" = "Turkey", "13" = "Australia",
  "14" = "New Zealand", "15" = "Canada",
  "16" = "U.S.A", "17" = "China/Hong Kong",
  "18" = "India", "19" = "Pakistan",
  "20" = "Bangladesh", "21" = "Sri Lanka",
  "23" = "Ghana", "24" = "Nigeria",
  "25" = "Uganda", "26" = "South Africa",
  "27" = "Jamaica", "97" = "Other country"
)

data <- data |>
  mutate(BirthCountry = ifelse(plbornc %in% names(country_map), 
                               country_map[as.character(plbornc)], NA)) 

data <- data |>
  mutate(generation = case_when(
    generation == 1 ~ "First Generation",
    generation == 2 ~ "Second Generation",
    generation %in% c(3, 4, 5, 6) ~ "Third Generation and Beyond",
    generation == -9 ~ NA_character_,
    TRUE ~ NA_character_  
  )) 

data <- data |>
  mutate(age_category = case_when(
    m_age_dv >= 18 & m_age_dv <= 24 ~ "18-24",
    m_age_dv >= 25 & m_age_dv <= 34 ~ "25-34",
    m_age_dv >= 35 & m_age_dv <= 44 ~ "35-44",
    m_age_dv >= 45 & m_age_dv <= 55 ~ "45-55",
    m_age_dv >= 56 & m_age_dv <= 64 ~ "56-64",
    m_age_dv >= 65 ~ "65+",
    TRUE ~ NA_character_  
  ))

3 Comparing Universal Credit Receipt between Native-Born and Immigrant Populations

proportions <- data |>
  filter(!is.na(UCrec) & !is.na(immigrant)) %>%
  group_by(immigrant) |>
  summarize(
    total = n(),
    uc_count = sum(UCrec == 1),
    proportion_UC = uc_count / total * 100,
    lower_ci = prop.test(uc_count, total)$conf.int[1] * 100,
    upper_ci = prop.test(uc_count, total)$conf.int[2] * 100
  )

total_individuals <- sum(!is.na(data$UCrec) & !is.na(data$immigrant))
total_receiving_uc <- sum(data$UCrec == 1 & !is.na(data$UCrec) & !is.na(data$immigrant))
proportion_uc <- total_receiving_uc / total_individuals
ci <- binom.test(total_receiving_uc, total_individuals)$conf.int
lower_ci <- ci[1] * 100
upper_ci <- ci[2] * 100
total_proportion <- data.frame(
  immigrant = "Total",
  total = total_individuals,
  uc_count = total_receiving_uc,
  proportion_UC = proportion_uc * 100,
  lower_ci = lower_ci,
  upper_ci = upper_ci
)
proportions <- bind_rows(proportions, total_proportion)

proportions <- proportions |>
  mutate(across(c(proportion_UC, lower_ci, upper_ci), ~sprintf("%.2f%%", .)))

proportions |>
  kable(format = "html", 
        caption = "Proportion of Individuals Receiving Universal Credit by Immigration Background",
        align = "c",
        col.names = c("Nationality Status", "Frequency", 
                      "Total Receiving UC", "Proportion Receiving UC (%)", "Lower CI (%)", "Upper CI (%)")) |>
  kable_styling(full_width = FALSE, font_size = 18)
Proportion of Individuals Receiving Universal Credit by Immigration Background
Nationality Status Frequency Total Receiving UC Proportion Receiving UC (%) Lower CI (%) Upper CI (%)
Immigrant 3399 269 7.91% 7.04% 8.89%
Native Born 23688 1195 5.04% 4.77% 5.33%
Total 27087 1464 5.40% 5.14% 5.68%

The initial findings from wave 13 of UKHLS indicate a higher likelihood of immigrants receiving UC compared to native-born individuals. The proportion of immigrants receiving UC was approximately 2.87% higher than the proportion of native-born individuals.

4 Analyzing Universal Credit Receipt across Generational Categories

proportions_generation <- data |>
  filter(!is.na(UCrec) & !is.na(generation)) |>
  group_by(generation) |>
  summarize(
    total_individuals = n(),
    total_UC_recipients = sum(UCrec == 1, na.rm = TRUE), 
    proportion_UC = round((total_UC_recipients / total_individuals) * 100, 2), 
    lower_ci = round(binom.test(total_UC_recipients, total_individuals)$conf.int[1] * 100, 2),
    upper_ci = round(binom.test(total_UC_recipients, total_individuals)$conf.int[2] * 100, 2) 
  )

proportions_generation |>
  kable(format = "html",
        caption = "Proportion of Individuals Receiving Universal Credit by Generation",
        align = "c",
        col.names = c("Generation", "Frequency", 
                      "Total Receiving UC", "Proportion Receiving UC (%)", "Lower CI (%)", "Upper CI (%)")) %>%
  kable_styling(full_width = FALSE, font_size = 18)
Proportion of Individuals Receiving Universal Credit by Generation
Generation Frequency Total Receiving UC Proportion Receiving UC (%) Lower CI (%) Upper CI (%)
First Generation 3397 269 7.92 7.03 8.88
Second Generation 2674 187 6.99 6.06 8.03
Third Generation and Beyond 19905 942 4.73 4.44 5.04

Generational background in the UK has a significant effect on the observed proportion of individuals receiving UC. In wave 13 of UKHLS, the proportion of first-generation immigrants receiving UC is approximately 0.93% higher than second generation immigrants (individuals with at least one parent born outside the UK). Moreover, the proportion of second-generation immigrants receiving UC is approximately 2.26% higher than individuals who are third generation (individuals whose parents are also born inside the UK) and beyond. Saliently, the proportion of first-generation immigrants receiving UC is approximately 3.19% higher than third or beyond generation individuals. First-generation immigrants are therefore the most likely recipients and third or beyond generation individuals are the least likely. # Comparison of Universal Credit Recipients by Country of Birth

proportion_UC_by_country <- data |>
  filter(!is.na(BirthCountry) & !is.na(UCrec)) |>
  group_by(BirthCountry) |>
  summarize(
    total_people = n(),
    total_receiving_credit = sum(UCrec == 1),
    prop_UC = mean(UCrec) * 100
  ) 

  ggplot(proportion_UC_by_country, aes(x = BirthCountry, y = prop_UC)) +
  geom_bar(stat = "identity", fill = "#5DA5DA", color = "black", width = 0.7) +
  geom_text(aes(label = paste0(round(prop_UC, 2), "%")), hjust = -0.2, size = 3, color = "black") +
  labs(x = "Country of Birth", y = "Proportion Receiving Universal Credit (%)") +
  coord_flip() +
  theme_minimal() +
  theme(axis.text.y = element_text(size = 10),
        axis.title.x = element_text(size = 11, face = "bold"),
        axis.title.y = element_text(size = 11, face = "bold"),
        plot.title = element_text(size = 13, face = "bold", hjust = 0.5),
        panel.grid.major.y = element_blank(),
        panel.grid.minor.y = element_blank(),
        panel.border = element_blank(),
        axis.line = element_line(color = "black")) +
  scale_y_continuous(limits = c(0, 20))

However, ‘immigrant’ is a capacious category vis-à-vis countries of origin, and “social welfare claims vary considerably by immigrant group”(Drinkwater & Robinson, 2011, p 24). Researchers have consistently found certain immigrant populations are more likely to claim social benefits than others. Consulting the bar chart visualizing the proportion of individuals receiving UC across countries of birth outside the UK, Wave 13 of UKLHS reflects these stark discrepancies. Individuals born in Bangladesh, Uganda, Nigeria, Pakistan and Turkey had proportions of UC receipt exceeding 10%. In contrast, there were no individuals born in Australia, Canada or New Zealand that mentioned claiming UC. Therefore, certain immigrant populations were far less likely to be the recipients of UC than the native-born British population and some were far more likely.

5 Universal Credit Receipt by Gender and Age among Immigrant and Native-Born Populations

filtered_data2 <- data %>%
  filter(!is.na(UCrec), !is.na(immigrant), !is.na(sex_recoded), !is.na(age_category))

grouped_data <- filtered_data2 %>%
  group_by(immigrant, sex_recoded, age_category)

summary_data <- grouped_data |>
  summarize(
    total_individuals = n(),
    total_receiving_uc = sum(UCrec == 1),
    proportion_UC = paste0(round(total_receiving_uc / total_individuals * 100, 2), "%"),
    .groups = "drop"  # Specify the `.groups` argument to drop the grouping
  )
    
summary_data |>
  kable(format = "html", 
        caption = "Summary of Proportion Receiving Universal Credit by Immigrant Status, Gender, and Age Category",
        align = "c",
        col.names = c("Immigrant Status", "Gender", "Age Category", 
                      "Total Individuals", "Total Receiving UC", "Proportion Receiving UC (%)")) %>%
  kable_styling(full_width = FALSE)
Summary of Proportion Receiving Universal Credit by Immigrant Status, Gender, and Age Category
Immigrant Status Gender Age Category Total Individuals Total Receiving UC Proportion Receiving UC (%)
Immigrant female 18-24 73 6 8.22%
Immigrant female 25-34 207 27 13.04%
Immigrant female 35-44 463 55 11.88%
Immigrant female 45-55 506 52 10.28%
Immigrant female 56-64 311 27 8.68%
Immigrant female 65+ 415 11 2.65%
Immigrant male 18-24 63 4 6.35%
Immigrant male 25-34 111 12 10.81%
Immigrant male 35-44 258 15 5.81%
Immigrant male 45-55 374 30 8.02%
Immigrant male 56-64 272 19 6.99%
Immigrant male 65+ 316 11 3.48%
Native Born female 18-24 1211 112 9.25%
Native Born female 25-34 1610 207 12.86%
Native Born female 35-44 1756 173 9.85%
Native Born female 45-55 2465 162 6.57%
Native Born female 56-64 2217 101 4.56%
Native Born female 65+ 3587 26 0.72%
Native Born male 18-24 990 78 7.88%
Native Born male 25-34 1196 91 7.61%
Native Born male 35-44 1349 81 6%
Native Born male 45-55 1985 84 4.23%
Native Born male 56-64 1794 60 3.34%
Native Born male 65+ 2985 15 0.5%

The age and gender composition of immigrant populations interact to influence the likelihood of UC receipt(Frattini & Dustmann,2014).In wave 13 of UKHLS, the proportion of individuals receiving UC for immigrant individuals ranges from 2.65% to 13.04% across these demographic groups, and 0.5% to 12.86% for native-born individuals. The proportion of UC recipients among males is generally lower for immigrants and native-born Britons. Saliently, immigrant females aged 25-34 had the highest proportion of individuals receiving UC followed by immigrant females aged 35-44. This further demonstrates the highest observed proportions of UC receipt are in younger and middle age categories. Hence, demographic variables like age and gender may account for the some of the difference between the proportions of immigrants and native-born individuals receiving UC.However, whilst these findings demonstrate complex interactions between nationality status, gender, age and proportion of individuals receiving UC: immigrant individuals generally exhibit higher proportions of UC receipt than native born individuals across most age and gender categories. Male and female immigrants, particularly in the age brackets of 25-34 and 35-44, are observed with higher proportions of individuals receiving UC. For instance, immigrant women age 25-34 have a approximately 3.79% higher proportion of individuals receiving UC than native-born women aged 25-34. Therefore, the pattern suggests immigrants, especially female immigrants, are more likely to rely on UC during their prime working years.

6 Conclusion and Discussion

Main findings: In wave 13 of UKHLS - Immigrants had an overall higher proportion of individuals receiving UC.
- First and second generation immigrants had a higher proportion of individuals receiving UC than third or beyond generation individuals. Hence, individuals with native-born parents seem to be less likely to receive UC than first and second generation immigrants. - There were vast discrepancies in the proportion of individuals receiving UC between immigrants born in different countries. Some immigrant groups had no UC claimants. Hence, not all immigrant groups are more likely to receive UC. Certain immigrants groups had large proportions of individuals, above 10%, receiving UC. Therefore, discussion and policymaking should perhaps focus on immigrants born in these countries. - Immigrant individuals generally exhibit higher proportions of UC receipt compared to native-born individuals across most age categories. Therefore, the impact of nationality status on the likelihood of UC receipt can not be explained away by other demographic variables, and the results point towards a multiplicative effect. - Immigrant women in their prime working years exhibit the highest likelihood of receiving UC.

It is important to recognise this paper solely employed descriptive analysis. Future research should incorporate logistical regression models to capture the effect of nationality status on UC receipt, whilst controlling for country of origin, age and gender. Therefore, conclusions on the validity or accurateness of ‘benefit tourism’ or ‘welfare dependence’ can not be extrapolated from this paper.

Word Count: 1352

7 Bibliography

Data source: Institute-for-Social-and-Economic-Research. (2023). Understanding Society: Waves 1-13, 2009-2022 and Harmonised BHPS: Waves 1-18, 1991-2009. University of Essex.

Anderson, B., & Blinder, S. (2024, February 23). Who Counts as a Migrant? Definitions and their Consequences. Retrieved from The Migration Observatory: At the University of Oxford: https://migrationobservatory.ox.ac.uk/resources/briefings/who-counts-as-a-migrant-definitions-and-their-consequences/

Bertolini, A., & Clegg , D. (2020). Access to Social Protection by Immigrants, Emigrants and Resident Nationals in the UK. Migration and Social Protection in Europe and Beyond (Volume 3):IMISCOE Research Series ((IMIS)), 419–432.

Drinkwater, S., & Robinson, C. (2011). Welfare Participation by Immigrants in the UK. International Journal of Manpower 34(2): DOI:10.1108/01437721311320636.

Frattini, T., & Dustmann , C. (2014). The Fiscal Effects of Immigration to the UK. Retrieved from UCL Department of Economics: https://www.ucl.ac.uk/economics/about-department/fiscal-effects-immigration-uk#:~:text=These%20differences%20are%20partly%20attributable,than%20natives%20to%20receive%20benefits.

Keen, R., & Apostolova , V. (2017). Statistics on migrants and benefits. House of Commons Library: Briefing Paper Number CBP 7445.

Mason, R. (2014, July 29). EU migrants to Britain face further restrictions on welfare payments. Retrieved from The Guardian: https://www.theguardian.com/uk-news/2014/jul/29/eu-migrants-britain-further-restrictions-welfare-payments.

National-Audit-Office. (2024). Progress in implementing Universal Credit. HC 552: Department for Work and Pensions.

Vargas-Silva, C., Sumption, M., & Walsh, W. P. (2022, march 30). The Fiscal Impact of Immigration in the UK. Retrieved from The Migration Observatory: At the University of Oxford : https://migrationobservatory.ox.ac.uk/resources/briefings/the-fiscal-impact-of-immigration-in-the uk/#:~:text=By%20contrast%2C%20using%20a%20similar,and%20£15.6bn%20respectively).