Revisiting the Integration Paradox: Higher Education and Perceived Discrimination Among Jews

IAB Colloquium Labour Market and Occupational Research

Christian Czymara (NIDI), Frank van Tubgergen (NIDI, UU)

12 Feb 2026

Introduction

  • Antisemitic incidents remain frequent across Europe, including verbal harassment, vandalism, physical attacks
  • Jews are, on average, more educated and socioeconomically well integrated, yet are often the target of hate crimes (RIAS 2024) and regularly report discrimination (Beyer and Liebe 2020; Enstad 2025)
  • Is there an “integration paradox” among Jews?

Theory

The Integration Paradox

  • Among majority, higher education is usually linked to tolerance and liberal values (Hainmueller and Hiscox 2007; Velásquez and Eger 2022)
  • For minorities, classic assimilation theory predicts that education facilitates integration into the host society (Gordon 1964)
  • However, research on migrants indicates that better access to mainstream society often leads to more discrimination (Steinmann 2019; Schaeffer & Kas 2023), known as the integration paradox
  • Thus, migrants with higher education often tend to perceive more discrimination (van Tubergen 2025)
  • Similar patterns for women (Lavest et al. 2025)
  • What about other minorities?

An “Integration Paradox” for Jews?

  • Evidence so far mostly focuses on migrants and ethno-racial minorities (and women)
  • Little is known about Jews as a religious minority
  • On the one side, highly educated Jews, like highly educated migrants, may interact more frequently with majority and have more awareness of bias
  • Hypothesis 1a: Higher education is associated with higher levels of perceived religious discrimination among Jews

An “Integration Paradox” for Jews?

  • On the other hand, Jews differ from (many) migrants in group size, visibility, integration
  • If highly educated Jews primarily interact with individuals with a similar background, they may be less likely to perceive discrimination
  • Hypothesis 1b: Higher education is associated with lower levels of perceived religious discrimination among Jews

Individual Religiosity

  • The association between education and personal religious discrimination should mostly concern Jews, for whom religion is important
  • For religious Jews, religion is more salient in daily life, and negative encounters are more likely to be interpreted as religion-based discrimination
  • For non-religious Jews, there are fewer religion-related interactions and a lower likelihood of attributing experiences to religious discrimination
  • Hypothesis 2a: The association between education and discrimination is stronger among religious Jews
  • Hypothesis 2b: The association between education and discrimination is stronger among those with a strong Jewish identity

Individual Visibility

  • Research on migrants shows that the integration paradox primarily exists among visible minorities (Schaeffer and Kas 2024)
  • Visibility increases likelihood of being identified as a minority and exposure to discriminatory behavior (Diehl and Trittler 2025; Flores 2015; van Tubergen and Kros 2025)
  • Similarly, education likely matters most for Jews who are visibly identifiable in public (e.g., through religious clothing or symbols)
  • Non-visible Jews likely face fewer religion-based interactions and thus lower risk of religious discrimination, regardless of education
  • Hypothesis 3: The association between education and perceived religious discrimination is stronger among visibly Jewish individuals

Intersection with Other Grounds of Discrimination

  • Religion is only one dimension of identity for Jews and different grounds of discrimination can overlap
  • Depending on their gender, migration background, or age, Jews may also face discrimination based on these grounds
  • As these are often visible characteristics, the integration paradox argument should hold
  • Hypothesis 4a: Positive education effect on ethnic/migration-based discrimination among Jews with a migration background
  • Hypothesis 4b: Positive education effect on gender-based discrimination among female Jews
  • Hypothesis 4c: Positive education effect on age-based discrimination among older Jews

Data

Second Survey on Discrimination and Hate Crime Against Jews

  • Collected European Union Agency for Fundamental Rights (FRA)
  • Targeting self-identified Jewish adults (16+) residing in 12 EU member states
  • Countries cover the vast majority of the EU’s Jewish population
  • Online opt-in survey design
  • Recruitment via multiple channels (Jewish communal organizations, Jewish and general media, social networks)

Outcome: Personal Perceived Religious Discrimination

  • FRA: In the PAST 12 MONTHS have you personally felt discriminated against in [COUNTRY] for any of the following reasons? –> Religion or belief (dummy: yes/no)
  • PEW: A combination of the following items (dummy if at least one is “yes”)
    • Have you been called offensive names because you are Jewish/of your Jewish background?
    • Have you been made to feel unwelcome because you are Jewish/of your Jewish background?
    • Have you been physically threatened or attacked because you are Jewish/of your Jewish background?
  • Discrimination in FRA data: 78, 22%, in the Pew data: 76, 24%

Predictor: Higher education

  • Dummy 1: tertiary vs. 0: less than tertiary
    • FRA: first and advanced level (ISCED 5, 6, 7, 8) = 1, else = 0
    • PEW: College graduate or postgrad = 1, else = 0
  • In FRA data, 71% hold higher education, in the Pew data: 81%

Method

  • Linear probability models with country fixed effects (same results with logistic regression or hierarchical linear models)
  • Regressing perceived discrimination on higher educational attainment
  • Controls:
    • Demographic characteristics (confounders): gender, age, marital status, migration background
    • Economic characteristics (colliders?): income (satisfaction), employment status
    • Religiousity (mediator)

Results

Impact of Education

  Bivariate Demographic controls Economic controls Religiosity
Predictors Estimates Estimates Estimates Estimates
high edu [Tertiary
education]
-0.05 ***
(0.01)
-0.04 ***
(0.01)
-0.02 **
(0.01)
-0.01
(0.01)
religiosity 0.03 ***
(0.00)
Observations 14036 14036 14036 14036
* p<0.05   ** p<0.01   *** p<0.001

Religiosity

  • Does the education effect depend on Jews’ level of religiosity?
  • FRA: How religious would you say you are? Please position yourself on a scale ranging from 1 Not at all religious to 10 Very religious
  • Controlling for confounders (male, age, migrant status), omitting mediators/colliders (income, employment status, marriage, religiosity), country FEs

Religiosity

  Religiosity Interaction
Predictors Estimates
high edu [Tertiary
education]
0.00
(0.01)
religiosity 0.03 ***
(0.00)
high edu [Tertiary
education] × religiosity
-0.01 ***
(0.00)
Observations 14036
* p<0.05   ** p<0.01   *** p<0.001

Religiosity

Identity

  • Does the education effect depend of respondents’ Jewish identity?
  • Interacting identity and education
  • FRA: Please position yourself on a scale ranging from 1 to 10 according to the strength of your Jewish identity, where 1 means very low strength and 10 means very high strength. (1 Very low strength - 10 Very high strength)
  • Same controls

Identity

  Identity Interaction
Predictors Estimates
high edu [Tertiary
education]
0.03
(0.03)
jident 0.03 ***
(0.00)
high edu [Tertiary
education] × jident
-0.01 **
(0.00)
Observations 14036
* p<0.05   ** p<0.01   *** p<0.001

Identity

Visibility

  • Does the education effect differ between those who are visibly Jewish vs those who are not?
  • Interacting visibility and education
  • FRA: Do you wear, carry or display things that might help people recognise you as a Jew in public [INFO BUTTON For example wearing a kippa/skullcap, magen david/Star of David or specific clothing, or displaying a mezuza]?
  • Same controls

Visibility

  Visibility Interaction
Predictors Estimates
high edu [Tertiary
education]
-0.05 ***
(0.01)
visible [Do not wear
anything distinctive]
-0.10 ***
(0.01)
high edu [Tertiary
education] × visible [Do
not wear anything
distinctive]
0.03
(0.01)
Observations 14036
* p<0.05   ** p<0.01   *** p<0.001

Visibility

The US

Jewish Americans in 2020 study

  • Large, nationally representative survey of Jewish adults in the US, conducted by Pew Research Center (2021) from November 2019 – June 2020
  • Probability-based design with online and mail modes
  • Stratified random sample of residential addresses of all 50 states + Washington, DC
  • Large initial screening survey identifying eligible respondents: Jews by religion and by identity, parentage, or upbringing
  • FRA survey is cross-national, including countries with varying histories of antisemitism
  • Pew survey is a probability sample in a context of long-standing Jewish inclusion

Impact of Education

  Bivariate Demographic controls Economic controls Religiosity
Predictors Estimates Estimates Estimates Estimates
high edu [Tertiary
education]
-0.06 ***
(0.02)
-0.06 ***
(0.02)
-0.04 *
(0.02)
-0.03
(0.02)
religiosity 0.06 ***
(0.01)
Observations 4867 4867 4867 4867
* p<0.05   ** p<0.01   *** p<0.001

Religiosity

  • PEW: How important is religion in your life? (4: Very important)
  • Same controls

Religiosity

  Religiosity Interaction
Predictors Estimates
high edu [Tertiary
education]
-0.03
(0.03)
religiosity 0.06 ***
(0.01)
high edu [Tertiary
education] × religiosity
-0.00
(0.01)
Observations 4867
* p<0.05   ** p<0.01   *** p<0.001

Religiosity

Visibility

  • PEW: On a typical day, do you wear something that is distinctively Jewish, such as a kippa (yarmulke) or a chain or bracelet with a Star of David or chai symbol?
  • Same controls

Visibility

  Visibility Interaction
Predictors Estimates
high edu [Tertiary
education]
-0.10 **
(0.03)
visible [Do not wear
anything distinctive]
-0.25 ***
(0.03)
high edu [Tertiary
education] × visible [Do
not wear anything
distinctive]
0.09 *
(0.04)
Observations 4867
* p<0.05   ** p<0.01   *** p<0.001

Visibility

Other Discrimination Types

Other Types of Discrimination among Subpopulations

  • What about the intersection with other “at risk characteristics”?
  • FRA data also asks about other grounds of discrimination
    • “Ethnic origin or immigrant background” discrimination for migrants only
    • “Age” discrimination for 70 or older only
    • “Sex/Gender” discrimination for women only
  • First, controling only demographic confounders, second including the full set of control variables

Other Types of Discrimination among Subpopulations

  Ethnic Discrimination among Migrants (demographic controls) Ethnic Discrimination among Migrants (full model) Age Discrimination among 70+ (demographic controls) Age Discrimination among 70+ (full model) Gender Discrimination among Women (demographic controls) Gender Discrimination among Women (full model)
Predictors Estimates Estimates Estimates Estimates Estimates Estimates
high_edu: Tertiary
education
0.02
(0.02)
0.04 *
(0.02)
0.05 **
(0.01)
0.06 ***
(0.01)
0.07 ***
(0.01)
0.08 ***
(0.01)
religiosity 0.01 **
(0.00)
-0.00
(0.00)
-0.01 ***
(0.00)
Observations 3636 3636 2863 2863 6809 6809
* p<0.05   ** p<0.01   *** p<0.001

Summing up

Summary & Conclusion

  • Contrary to the integration paradox, higher education is associated with lower perceived discrimination among Jews
  • This contrasts with findings for migrants and women (even in this sample)
  • Jews are often not readily identifiable in everyday interactions
  • Yet, education reduces perceived discrimination more strongly among visibly Jewish individuals
  • Visibility interacts with education in a protective rather than amplifying way
  • A paradox in the paradox?
  • Education may enable exposure to socially or institutionally protected contexts, where visible Jewishness is more accepted

Annex

Number of Respondents by Country


    Belgium     Denmark     Germany       Spain      France       Italy 
        648         506         986         475        3335         611 
     Latvia     Hungary Netherlands     Austria      Poland      Sweden 
          0         534        1053         442         351        1016 
         UK 
       4079 

Distribution across Countries