Pieter-Paul Verhaeghe
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Pieter-Paul Verhaeghe has a total of 70 co-authors. Here are the top co-authors by number of shared publications:
Bart Van de Putte
Department of Sociology Ghent University Ghent Belgium
10 shared publications- Abel Ghekiere
Department of Sociology, Vrije Universiteit Brussel, Brussels, Belgium
8 shared publications - Koen Van der Bracht
Ghent University, Belgium
5 shared publications - Billie Martiniello
Department of Sociology, Vrije Universiteit Brussel, Brussels, Belgium
4 shared publications - Dounia Bourabain
School of Social Sciences, Hasselt University
4 shared publications - Ad Coenen
Department of Sociology Ghent University Ghent Belgium
3 shared publications
Latest Publications
- 13 October 2026
- Published by Elsevier in Social Science & Medicine
- Vol. 410
Unemployment and mental health: Evidence of effect heterogeneity by social class in the UK
- 13 October 2026
- Published by Elsevier in Social Science & Medicine
- Vol. 410
Unemployment and mental health: Evidence of effect heterogeneity by social class in the UK
- Open Access
- 9 September 2026
- Published by Center for Open Science in SocArXiv
From 4-Digit Codes to Social Class: Benchmarking Supervised Classification, Semantic Retrieval, and LLM Reranking for Multilingual Occupational Coding and Downstream Stratification (Technical report: model design, evaluation, and derived-measure accuracy)
This report documents the design, implementation, and evaluation of a fully local pipeline that codes multilingual free-text occupation responses into the International Standard Classification of Occupations (ISCO-08) and derives from those codes the social-class and status measures (e.g., ESeC, ISEI, Oesch) on which much of stratification research depends. This coding is costly, slow, and only moderately consistent when done by hand. The pipeline is local (hence GDPR-friendly), reproducible, and runs on consumer-grade hardware. We benchmark and compare three approaches to the coding step: (M1) a supervised multilingual transformer classifier; (M2) semantic retrieval against a 361k-document multilingual occupation index; and (M3) the supervised classifier with an added large-language-model reranking step. On held-out European Social Survey (ESS) data the supervised classifier reaches 59.2% exact four-digit accuracy, comparable to human inter-coder agreement for this intrinsically ambiguous task, which is itself low at the four-digit level (reported by past research to be ≈42–53% between individual coders; κ ≈ 0.5 between agencies). More importantly for stratification research, given that coding errors are mostly near-misses that coarser schemes absorb, the sociologically relevant aggregate measures are recovered far more accurately, namely 80.4% for the 9-class ESeC (87.8% for its 5-class collapse), 80.5% for the 7-group ESeG, 77.9% for the 16-category Oesch scheme, and a Pearson correlation of 0.87 for the ISEI socio-economic index. The same class- and status-level accuracy holds when the model is trained and tested on canonical occupation text, so it is not an artefact of the ESS sample. Retrieval alone (M2) is substantially weaker, and contrary to the intuitive expectation, adding an LLM reranker (M3) reduces accuracy on noisy multilingual survey responses (−3.7 pp), while on clean canonical titles its effect is approximately neutral, i.e., +0.5 pp over the backbone it reranks, with no gain over the best supervised model. Instead, the supervised model’s (M1) well-calibrated confidence (ECE = 0.038) supports a human-in-the-loop workflow in which only genuinely uncertain cases are escalated to human coders. We report bootstrap confidence intervals and paired significance tests for every comparison, a train/test leakage audit, and a decomposition of geographic variation showing that the cross-country accuracy spread (in the ESS corpus) reflects response style and occupational-title ambiguity rather than language resources, whereas a genuine resource effect appears only across languages on canonical text in the normative corpus. The pipeline runs entirely offline on a single consumer-grade workstation, making automated, auditable occupation coding feasible without sending respondent data to external services.
- 9 September 2026
- Published by Center for Open Science in SocArXiv
Open AccessFrom 4-Digit Codes to Social Class: Benchmarking Supervised Classification, Semantic Retrieval, and LLM Reranking for Multilingual Occupational Coding and Downstream Stratification (Technical report: model design, evaluation, and derived-measure accuracy)
This report documents the design, implementation, and evaluation of a fully local pipeline that codes multilingual free-text occupation responses into the International Standard Classification of Occupations (ISCO-08) and derives from those codes the social-class and status measures (e.g., ESeC, ISEI, Oesch) on which much of stratification research depends. This coding is costly, slow, and only moderately consistent when done by hand. The pipeline is local (hence GDPR-friendly), reproducible, and runs on consumer-grade hardware. We benchmark and compare three approaches to the coding step: (M1) a supervised multilingual transformer classifier; (M2) semantic retrieval against a 361k-document multilingual occupation index; and (M3) the supervised classifier with an added large-language-model reranking step. On held-out European Social Survey (ESS) data the supervised classifier reaches 59.2% exact four-digit accuracy, comparable to human inter-coder agreement for this intrinsically ambiguous task, which is itself low at the four-digit level (reported by past research to be ≈42–53% between individual coders; κ ≈ 0.5 between agencies). More importantly for stratification research, given that coding errors are mostly near-misses that coarser schemes absorb, the sociologically relevant aggregate measures are recovered far more accurately, namely 80.4% for the 9-class ESeC (87.8% for its 5-class collapse), 80.5% for the 7-group ESeG, 77.9% for the 16-category Oesch scheme, and a Pearson correlation of 0.87 for the ISEI socio-economic index. The same class- and status-level accuracy holds when the model is trained and tested on canonical occupation text, so it is not an artefact of the ESS sample. Retrieval alone (M2) is substantially weaker, and contrary to the intuitive expectation, adding an LLM reranker (M3) reduces accuracy on noisy multilingual survey responses (−3.7 pp), while on clean canonical titles its effect is approximately neutral, i.e., +0.5 pp over the backbone it reranks, with no gain over the best supervised model. Instead, the supervised model’s (M1) well-calibrated confidence (ECE = 0.038) supports a human-in-the-loop workflow in which only genuinely uncertain cases are escalated to human coders. We report bootstrap confidence intervals and paired significance tests for every comparison, a train/test leakage audit, and a decomposition of geographic variation showing that the cross-country accuracy spread (in the ESS corpus) reflects response style and occupational-title ambiguity rather than language resources, whereas a genuine resource effect appears only across languages on canonical text in the normative corpus. The pipeline runs entirely offline on a single consumer-grade workstation, making automated, auditable occupation coding feasible without sending respondent data to external services.
- 10 August 2026
- Published by Taylor & Francis in Ethnic and Racial Studies
- p. 1-21
Discrimination and deservingness: public attitudes towards ethnic, linguistic, and religious exclusion in Belgium
Discrimination based on various characteristics and in real-world settings like labour and housing markets continues to challenge equality and inclusivity in societies worldwide (Del Carmen Triana ...
- 10 August 2026
- Published by Taylor & Francis in Ethnic and Racial Studies
- p. 1-21
Discrimination and deservingness: public attitudes towards ethnic, linguistic, and religious exclusion in Belgium
Discrimination based on various characteristics and in real-world settings like labour and housing markets continues to challenge equality and inclusivity in societies worldwide (Del Carmen Triana ...
- Open Access
- 4 July 2026
- Published by Springer Nature in De Economist
The Relative Rates of Ethnic, Age and Gender Discrimination and Their Intersectionality: A Factorial Field Experiment in the Brussels Capital Region
This study investigates ethnic, age, and gender discrimination, and their intersectionality in Brussels’ ethnically highly diverse metropolitan labour market. Using a factorial correspondence experiment, we assess the causal impact of ethnicity (Polish or Moroccan vs. Belgian-sounding names), age (6- or 12-year gap with applicants aged 38 or 44), and gender (male vs. female) on job interview invitations and any positive response. All applicants were born in Belgium, allowing us to isolate discrimination based on perceived ethnic origin rather than migration status. Findings indicate that applicants with a Moroccan (non-European) ethnic background, a 12-year age gap, and male gender face discrimination compared to those of Belgian ethnicity, younger age, and female gender respectively. Interestingly, the magnitude of these discriminatory effects is comparable across all three dimensions. No discrimination was observed between Polish and Belgian applicants and applicants with a 6-year age gap. Last, we find no robust evidence of intersectional effects between the tested characteristics, likely attributable to limited statistical power.
- 4 July 2026
- Published by Springer Nature in De Economist
Open AccessThe Relative Rates of Ethnic, Age and Gender Discrimination and Their Intersectionality: A Factorial Field Experiment in the Brussels Capital Region
This study investigates ethnic, age, and gender discrimination, and their intersectionality in Brussels’ ethnically highly diverse metropolitan labour market. Using a factorial correspondence experiment, we assess the causal impact of ethnicity (Polish or Moroccan vs. Belgian-sounding names), age (6- or 12-year gap with applicants aged 38 or 44), and gender (male vs. female) on job interview invitations and any positive response. All applicants were born in Belgium, allowing us to isolate discrimination based on perceived ethnic origin rather than migration status. Findings indicate that applicants with a Moroccan (non-European) ethnic background, a 12-year age gap, and male gender face discrimination compared to those of Belgian ethnicity, younger age, and female gender respectively. Interestingly, the magnitude of these discriminatory effects is comparable across all three dimensions. No discrimination was observed between Polish and Belgian applicants and applicants with a 6-year age gap. Last, we find no robust evidence of intersectional effects between the tested characteristics, likely attributable to limited statistical power.
- Open Access
- 28 November 2025
- Published by Springer Nature in Scientific Data
- Vol. 12 (1) , 1883
The Perception of Names in Experimental Studies on Ethnic Origin: A Cross-National Validation in Europe
A growing body of research uses field and survey experiments to examine ethnic discrimination. Central to these studies is the use of people’s names as a proxy for ethnic origin. However, names signal more than solely ethnic markers. Moreover, their signals might vary across national contexts. Scholars should pre-test the perception of names used in experiments in order to properly interpret their results and reveal the mechanisms underlying discrimination. There is, however, no comprehensive study yet in Europe which thoroughly pre-test the perception of names across countries with profoundly different migration histories. In this paper, we present the dataset ‘Perceptions of names in Europe’, containing the perceptions of 1078 names studied across nine European countries: Belgium, Czech Republic, Germany, Hungary, Ireland, the Netherlands, Spain, Switzerland, and the United Kingdom. The dataset includes 82.400 evaluations from 8.240 respondents about the distinctiveness of Sub-Sahara African, Muslim and Roma names in terms of minority-majority group status, gender, religiosity, socioeconomic status, skin colour, and language proficiency. Information on respondents’ background characteristics are also available.
- 28 November 2025
- Published by Springer Nature in Scientific Data
- Vol. 12 (1) , 1883
Open AccessThe Perception of Names in Experimental Studies on Ethnic Origin: A Cross-National Validation in Europe
A growing body of research uses field and survey experiments to examine ethnic discrimination. Central to these studies is the use of people’s names as a proxy for ethnic origin. However, names signal more than solely ethnic markers. Moreover, their signals might vary across national contexts. Scholars should pre-test the perception of names used in experiments in order to properly interpret their results and reveal the mechanisms underlying discrimination. There is, however, no comprehensive study yet in Europe which thoroughly pre-test the perception of names across countries with profoundly different migration histories. In this paper, we present the dataset ‘Perceptions of names in Europe’, containing the perceptions of 1078 names studied across nine European countries: Belgium, Czech Republic, Germany, Hungary, Ireland, the Netherlands, Spain, Switzerland, and the United Kingdom. The dataset includes 82.400 evaluations from 8.240 respondents about the distinctiveness of Sub-Sahara African, Muslim and Roma names in terms of minority-majority group status, gender, religiosity, socioeconomic status, skin colour, and language proficiency. Information on respondents’ background characteristics are also available.
- Open Access
- 27 November 2025
- Published by Wiley in Sociology Compass
- Vol. 19 (12)
Five Traditions of Quantitative Research on Racism: Their Research Objects, Methods, Assumptions and Relationships
The study of racism has become increasingly fragmented, with varying conceptualizations, specializations and methodologies, often divided along disciplinary lines. This paper reviews five traditions of quantitative research on racism, covering research on explicit racial attitudes, implicit biases, discrimination, migration and integration policies, ethno‐racial disparities, and perceived experiences of racism. We elaborate on their research objects, historical development, measurement methods, core assumptions and critiques, and main findings. Moreover, we discuss how these five traditions are interrelated and connected to the broader concept of racism at the micro‐, meso‐, and macro‐level. Our position is that quantitative research can be used for racial justice, but it requires deeper transdisciplinary collaboration across the five traditions, more critical reflection on their underlying assumptions and limitations, and more explicit theorizing about how they relate to the multifaceted nature of racism.
- 27 November 2025
- Published by Wiley in Sociology Compass
- Vol. 19 (12)
Open AccessFive Traditions of Quantitative Research on Racism: Their Research Objects, Methods, Assumptions and Relationships
The study of racism has become increasingly fragmented, with varying conceptualizations, specializations and methodologies, often divided along disciplinary lines. This paper reviews five traditions of quantitative research on racism, covering research on explicit racial attitudes, implicit biases, discrimination, migration and integration policies, ethno‐racial disparities, and perceived experiences of racism. We elaborate on their research objects, historical development, measurement methods, core assumptions and critiques, and main findings. Moreover, we discuss how these five traditions are interrelated and connected to the broader concept of racism at the micro‐, meso‐, and macro‐level. Our position is that quantitative research can be used for racial justice, but it requires deeper transdisciplinary collaboration across the five traditions, more critical reflection on their underlying assumptions and limitations, and more explicit theorizing about how they relate to the multifaceted nature of racism.
- 18 September 2025
- Published by Emerald Publishing in International Journal of Public Sector Management
- Vol. 39 (3) , 381-399
Even grounds: assessing discrimination in recruitment in the public sector
Purpose: Public sector employers often struggle to meet diversity targets, even with standardized processes designed to ensure fairness in recruitment. This raises a critical question: Is hiring discrimination a key driver of the continued underrepresentation of ethno-national minorities and individuals with different health statuses in the public sector? This study examines this question within the context of the Flemish regional government. Design/methodology/approach: The study centres on a duo- and trio-correspondence field experiment, submitting 437 fictitious applications to 207 job vacancies within the Flemish government. Matched applications were identical in qualifications but differed on signals of ethno-national origin (Moroccan or Polish vs Belgian), gender and health status (signalling an impairment vs none). Hiring discrimination is then measured by comparing the callback rates for invitations to the next selection round between these different profiles. Findings: No statistically significant discrimination based on ethno-national origin or health status was found in the CV-screening phase. However, the study uncovered a significant gender bias: male recruiters were significantly more likely to invite male applicants than female applicants (62.9% vs. 39.0%), suggesting the presence of in-group favouritism, or homophily, among male recruiters. Research limitations/implications: This research has two main limitations. First, the study’s focus on the initial CV screening phase is a limitation inherent to the method, as discrimination can occur at other recruitment stages. Second, the sample size was insufficiently large to conduct detailed intersectional analyses such as testing whether the observed gender bias varies by job type (e.g. in traditionally male- or female-dominated occupations) or interacts with ethno-national origin across different job levels. Practical implications: The findings, particularly the absence of hiring discrimination against ethno-national minorities and individuals with different health statuses, provide several implications. First, for Flemish public administrators, this study offers a data-driven directive: resources should be strategically redirected from combating hiring discrimination towards initiatives that attract more candidates from these specific communities and ensure their retention after hiring. Additionally, this study offers a crucial lesson for the private sector by demonstrating that the public sector’s standardized processes provide a successful model for mitigating ethnic- and health-based discrimination. Finally, in a context of declining public trust in government, the findings offer a crucial counter-narrative. They provide empirical evidence that the public sector’s initial hiring phase is procedurally fair for candidates, regardless of ethno-national origin or health status. Originality/value: This study provides a rare empirical test of hiring discrimination in the under-tested public sector, contrasting with the extensive focus on the private sector labour markets. It advances correspondence testing methodology by employing a job-category-agnostic approach, enhancing external validity beyond studies that typically target a limited set of occupations. The study also explores the less-studied ground of health-status discrimination.
- 18 September 2025
- Published by Emerald Publishing in International Journal of Public Sector Management
- Vol. 39 (3) , 381-399
Even grounds: assessing discrimination in recruitment in the public sector
Purpose: Public sector employers often struggle to meet diversity targets, even with standardized processes designed to ensure fairness in recruitment. This raises a critical question: Is hiring discrimination a key driver of the continued underrepresentation of ethno-national minorities and individuals with different health statuses in the public sector? This study examines this question within the context of the Flemish regional government. Design/methodology/approach: The study centres on a duo- and trio-correspondence field experiment, submitting 437 fictitious applications to 207 job vacancies within the Flemish government. Matched applications were identical in qualifications but differed on signals of ethno-national origin (Moroccan or Polish vs Belgian), gender and health status (signalling an impairment vs none). Hiring discrimination is then measured by comparing the callback rates for invitations to the next selection round between these different profiles. Findings: No statistically significant discrimination based on ethno-national origin or health status was found in the CV-screening phase. However, the study uncovered a significant gender bias: male recruiters were significantly more likely to invite male applicants than female applicants (62.9% vs. 39.0%), suggesting the presence of in-group favouritism, or homophily, among male recruiters. Research limitations/implications: This research has two main limitations. First, the study’s focus on the initial CV screening phase is a limitation inherent to the method, as discrimination can occur at other recruitment stages. Second, the sample size was insufficiently large to conduct detailed intersectional analyses such as testing whether the observed gender bias varies by job type (e.g. in traditionally male- or female-dominated occupations) or interacts with ethno-national origin across different job levels. Practical implications: The findings, particularly the absence of hiring discrimination against ethno-national minorities and individuals with different health statuses, provide several implications. First, for Flemish public administrators, this study offers a data-driven directive: resources should be strategically redirected from combating hiring discrimination towards initiatives that attract more candidates from these specific communities and ensure their retention after hiring. Additionally, this study offers a crucial lesson for the private sector by demonstrating that the public sector’s standardized processes provide a successful model for mitigating ethnic- and health-based discrimination. Finally, in a context of declining public trust in government, the findings offer a crucial counter-narrative. They provide empirical evidence that the public sector’s initial hiring phase is procedurally fair for candidates, regardless of ethno-national origin or health status. Originality/value: This study provides a rare empirical test of hiring discrimination in the under-tested public sector, contrasting with the extensive focus on the private sector labour markets. It advances correspondence testing methodology by employing a job-category-agnostic approach, enhancing external validity beyond studies that typically target a limited set of occupations. The study also explores the less-studied ground of health-status discrimination.
- Open Access
- 10 December 2024
- Published by Cambridge University Press (CUP) in The Journal of Race, Ethnicity, and Politics
- Vol. 10 (2) , 426-453
Local Governments’ Response to Discrimination: A Temporal Framework to Analyze Local Anti-discrimination Policy and Actions
Discrimination is widely studied, with extensive research measuring discrimination on the housing and labor markets. This study examines how local governments address this well-documented issue, by conducting content analysis on 45 policy documents and by performing semi-structured in-depth interviews with 24 alder(wo)men and diversity officers across nine Belgian cities. We introduce a temporal framework combining why, what, how, and when local anti-discrimination policy and actions are established. Such a framework is useful, as we do not approach policy as fixed, but pay attention to how actions evolve over time, even within one so-called anti-discrimination policy.This enables scholars and policymakers to identify decision-making patterns, predict changes over time, and understand contextual influences. Besides, unlike existing models rooted in integration or diversity policy, our framework captures the unique aspects of anti-discrimination policy, enabling a thorough understanding of the (non-)adoption of concrete anti-discrimination actions.
- 10 December 2024
- Published by Cambridge University Press (CUP) in The Journal of Race, Ethnicity, and Politics
- Vol. 10 (2) , 426-453
Open AccessLocal Governments’ Response to Discrimination: A Temporal Framework to Analyze Local Anti-discrimination Policy and Actions
Discrimination is widely studied, with extensive research measuring discrimination on the housing and labor markets. This study examines how local governments address this well-documented issue, by conducting content analysis on 45 policy documents and by performing semi-structured in-depth interviews with 24 alder(wo)men and diversity officers across nine Belgian cities. We introduce a temporal framework combining why, what, how, and when local anti-discrimination policy and actions are established. Such a framework is useful, as we do not approach policy as fixed, but pay attention to how actions evolve over time, even within one so-called anti-discrimination policy.This enables scholars and policymakers to identify decision-making patterns, predict changes over time, and understand contextual influences. Besides, unlike existing models rooted in integration or diversity policy, our framework captures the unique aspects of anti-discrimination policy, enabling a thorough understanding of the (non-)adoption of concrete anti-discrimination actions.
- Open Access
- 1 January 2024
- Published by SAGE Publications in Socius: Sociological Research for a Dynamic World
- Vol. 10
Implicit Racial Bias in Europe: Cross-National Variation and Time Trends
The authors visualize how implicit racial bias at the collective level varies across European countries and how it changed between 2009 and 2019. To obtain population-level estimates, the authors re-weight implicit association test data from the Project Implicit international dataset ( n = 184,745). Implicit bias against Black people is stronger in southern and eastern vis-à-vis northern and western European countries. It decreased until the mid-2010s but started to increase thereafter.
- 1 January 2024
- Published by SAGE Publications in Socius: Sociological Research for a Dynamic World
- Vol. 10
Open AccessImplicit Racial Bias in Europe: Cross-National Variation and Time Trends
The authors visualize how implicit racial bias at the collective level varies across European countries and how it changed between 2009 and 2019. To obtain population-level estimates, the authors re-weight implicit association test data from the Project Implicit international dataset ( n = 184,745). Implicit bias against Black people is stronger in southern and eastern vis-à-vis northern and western European countries. It decreased until the mid-2010s but started to increase thereafter.
- Open Access
- 27 February 2023
- Published by Taylor & Francis in Ethnic and Racial Studies
- Vol. 46 (12) , 2654-2676
Identifying rental discrimination on the Flemish housing market: an intersectional approach
This study aims to identify rental discrimination on the Flemish rental housing market in Belgium, taking the intersectional nature of discrimination into account. Most discrimination studies focus on unequal access based on one individual discrimination ground. This practice eludes the context of the other discrimination grounds in which rental discrimination occurs and neglects the intersectional nature of discrimination. We therefore conducted 8.245 correspondence tests in almost all municipalities in Flanders, the Dutch-speaking part of Belgium. We apply an intersectional lens, by considering the relation and interaction between gender (male/female), ethnic origin (Moroccan/Polish) and the homogeneity of names (homogenous/mixed). We find four strata of rental discrimination, in which Moroccan female rental candidates with a homogenous name experience most discrimination, indicating that multiple categorical identities enforce each other. Without a full intersectional approach, these layered patterns of exclusion would have been hidden behind the bold boundaries of ethnic or gender categorizations.
- 27 February 2023
- Published by Taylor & Francis in Ethnic and Racial Studies
- Vol. 46 (12) , 2654-2676
Open AccessIdentifying rental discrimination on the Flemish housing market: an intersectional approach
This study aims to identify rental discrimination on the Flemish rental housing market in Belgium, taking the intersectional nature of discrimination into account. Most discrimination studies focus on unequal access based on one individual discrimination ground. This practice eludes the context of the other discrimination grounds in which rental discrimination occurs and neglects the intersectional nature of discrimination. We therefore conducted 8.245 correspondence tests in almost all municipalities in Flanders, the Dutch-speaking part of Belgium. We apply an intersectional lens, by considering the relation and interaction between gender (male/female), ethnic origin (Moroccan/Polish) and the homogeneity of names (homogenous/mixed). We find four strata of rental discrimination, in which Moroccan female rental candidates with a homogenous name experience most discrimination, indicating that multiple categorical identities enforce each other. Without a full intersectional approach, these layered patterns of exclusion would have been hidden behind the bold boundaries of ethnic or gender categorizations.