Research

This paper examines the impact of delayed retirement, induced by pension reforms, on late-career mental health, focusing on working conditions. While studies have analysed aspects of job quality – such as highstrain roles and automation risk – none have considered the full range of job characteristics shaping workers’ experiences. We address this gap by analysing six key dimensions of job quality: skills and discretion, working time quality, physical environment, social environment, work intensity, and career prospects. To mitigate endogeneity concerns in self-reported mental health measures, we incorporate occupation-level data on working conditions from external sources. 

Our analysis leverages pension reforms enacted between 2011 and 2015 in 14 European countries, integrating data from the Survey of Health, Ageing and Retirement in Europe with job quality measures from the European Working Conditions Survey. Using a staggered difference-in-differences design, we estimate the causal impact of extended work horizons on depression while accounting for cross-country differences in labour markets and pension systems. 

Our findings confirm that delaying retirement negatively affects older workers’ mental health, especially in the case of larger increases in the retirement age. However, the magnitude of this impact varies depending on job quality. Workers in unsupportive social environments, precarious jobs with limited career prospects, or roles with low autonomy exhibit the largest increases in depression. In contrast, individuals in high quality jobs, particularly those with supportive workplaces, experience milder negative effects or even benefits. To prevent pension reforms from harming workers’ well-being, they should be complemented by labour market policies that promote sustainable working conditions, job adaptability, and lifelong learning.

New technologies for human augmentation have emerged over the years and changed the relationship between humans and technology from an interaction to an integration (e.g., exoskeletons in the workplace). Although the literature has discussed bodily integration and the impact of human augmentation technology on embodied experience, the specific type of integration and its effect on attitudes toward technology remain unclear. Therefore, this paper investigates the dimensions characterizing the relationship emerging from workers’ use of exoskeletons. We filmed workers performing tasks with and without exoskeletons. These workers then conducted two Q-sorts about their embodied experiences. By analyzing Q-data and video footage, we identified non-hierarchical models of human- technology relationships. We observed four bodily integration types to various degrees: super body (i.e., a high sense of agency and body ownership), chauffeured body (i.e., a low sense of agency and low sense of body ownership), tele- body (e., a high sense of agency and low sense of body ownership), and possessed body (i.e., a low sense of agency and a high sense of body ownership). We found a link between “super body” and a positive, accepting, attitude toward exoskeletons despite the related long-term ethical issues. We also found that participants who experienced a “chauffeured body” did so because they had difficulty regaining control over their movements. Therefore, strategies for developing exoskeletons should focus on personalized and transparent algorithms that manage human-exoskeleton interactions while increasing user agency from several angles.

Mental health issues among researchers and doctoral students have become an increasing concern, yet existing studies often provide only cross-sectional evidence from specific national contexts and offer limited insight into how academic communities themselves discuss mental health. This paper examines public conversations about mental health in academia by analyzing Twitter discussions within French and UK academic communities. Focusing on tweets related to PhD studies, stress, mental health, burnout, and depression, the study aims to identify key themes in these conversations, compare discussions before, during, and after the peak of the COVID-19 crisis, and explore differences between doctoral researchers, other academics, and academic institutions. By comparing France and the UK, the paper also considers how national policy contexts and public awareness initiatives may shape online discourse around mental health in academia. The analysis contributes to a better understanding of the concerns expressed by researchers and doctoral students, the role of social media as a space for discussing academic distress, and the potential use of such evidence to inform more targeted mental health policies and support programs for the academic community.

Automation and robotics are key enablers of Industry 4.0, as they include a promise of increased productivity through changes in work practices. However, they also have ethical issues, with values being challenged in the economic, social, political, and industrial realms. Thus, we used generative artificial intelligence to generate illustrations of these ethical dimensions, tested the pictures (n=20; n=15) and included them in a visual Q-study that was administered in four countries (Australia, South Africa, the United Kingdom, and the United States of America; n=90) to identify the dimensions that laypeople perceived as troubling. We analyzed the results in each country and then ran a second-order analysis to compare the views across countries. Three composite views, including an idiosyncratic view, emerged. The first view prioritizes controlling automation to prevent ecological and security risks. The second view values robots but cautions against close human–robot interactions. The third view focuses on concerns about changing work dynamics and inclusivity. Based on these results, we propose a framework for assessing attitudes toward automation technologies that includes variables such as sustainability, meaning of work, trust in technology and humans, usefulness, and economic impact on individuals. 

Highlight:

« Q methodology is a way to understand how people think about a topic. It is useful when you want to look closely at their opinions, values, or beliefs. For example, researchers have used it to study what older people think about well-being (Hackert et al., 2019), whether new technologies are ethical and acceptable (Gauttier, 2019), or political ideas in the United States (Leonard et al., 2021). In Q methodology, participants are given a set of statements and asked to arrange them to show which ones they agree with most and which ones they disagree with. Researchers then study these arrangements to find common viewpoints among participants. This helps show how and why people have different opinions about the same topic.» 

This paper examines whether gender pay inequality within firms affects employees’ mental well-being and through which mechanisms such effects may operate. Using French matched employer–employee data combined with administrative measures of firm-level gender pay gaps, I show that although women consistently report poorer mental well-being than men,the magnitude of the gender pay gap within firms is not economically meaningful in explaining this disparity. However, the association between inequality and well-being is heterogeneous: consistent with relative deprivation theory, lower-ranked employees — particularly women — experience a stronger negative relationship between pay inequality and well-being, whereas higher-ranked employees are less affected. I further examine the role of salience by exploiting the 2019 introduction of the Egapro Index, which mandated public disclosure of gender equality indicators for large firms in France. Using a difference-in-differences design, I find that disclosure generates sizeable short-run improvements in well-being among male employees, with no comparable gains for women. These results suggest that inequality does not affect well-being uniformly, but becomes consequential when it is made salient. Transparency policies can generate positive psychological effects, but without accompanying measures that reduce underlying pay disparities, their benefits may accrue unevenly across groups.

This paper estimates the causal effect of job insecurity on health and health-related behaviours among workers aged 50 and above. Major labour market downturns, such as the 2008 financial crisis, generated sharp and heterogeneous employment shocks across sectors and occupations, substantially increasing workers’ exposure to job insecurity. Older workers are particularly vulnerable to such shocks due to lower re-employment prospects and higher baseline health risks.

Using longitudinal data from the Survey of Health, Ageing and Retirement in Europe (SHARE) covering 19 European countries between 2004 and 2015, we implement an instrumental variable strategy that exploits exogenous variation in job insecurity induced by the Great Recession. Job insecurity is instrumented using a shift-share measure of occupation-specific exposure to industry-level employment shocks constructed from European Labour Force Survey (EU-LFS) data.

The results show that higher job insecurity reduces subjective well-being and self-assessed health, increases obesity and chronic conditions. At the same time, it lowers smoking and alcohol consumption. Instrumental variable estimates exceed ordinary least squares ones, indicating that conventional approaches underestimate the health costs of job insecurity. These findings suggest that policies aimed at stabilising employment expectations and supporting older workers during labour market downturns may yield substantial public health benefits.

This paper studies how mandatory gender equality disclosure affects worker mobility. We examine France’s Egapro reform, which required firms to publicly report gender equality indicators under a staggered implementation schedule based on firm size. Using linked employer–employee administrative data, we exploit variation in exposure across firm-size thresholds and implementation years to estimate whether transparency changes employee turnover between firms. The analysis asks whether workers respond to disclosed gender equality information by leaving lower-performing firms, and whether responses differ by gender, occupation, wage level, or local labor market conditions. The project contributes to evidence on pay transparency, workplace inequality, and the role of information in shaping labor market choices. Work in progress. 

Remote work has expanded rapidly with digitalisation and the COVID-19 shock, yet evidence on its consequences for workers is mixed. This paper studies how working from home (WFH) affects the balance between job effort and rewards, using the Effort–Reward Imbalance (ERI) framework, and whether effects differ by personality. To address selection into telework, we estimate causal effects with a two-stage least squares strategy that instruments individual WFH status with an occupation-level index of technical teleworkability matched at the three-digit ISCO-08 level. Work in progress. 

Work in progress.

This paper examines the use of wearable technologies to detect and interpret stress among hospital staff. Based on a field study with 11 employees in the geriatrics department of an Italian hospital, stress was measured using the Empatica E4 wristband, which collects physiological signals such as electrodermal activity, blood volume pulse, skin temperature, and movement data. These data were combined with self-reported stress measures, including the Perceived Stress Scale and the HSE work-related stress questionnaire. Several machine learning models were trained and compared for stress detection, with the k-nearest neighbors model achieving the best accuracy when activity data were included. The results show that detected stress episodes broadly correspond to employees’ perceived stress levels, although individuals tend to assess stress more by intensity than by frequency. The study also reveals differences in stress factors across professional groups: nurses report stress linked to lack of autonomy and organizational change, while doctors are more affected by workload and pressure. Overall, the wearable device appears useful for organizational understanding of workplace stress, though its added value for individuals’ own interpretation of stress remains limited.

Work in progress.

The rise of digital platforms has significantly reshaped how companies access external labor markets, blurring the boundaries between internal and external workforces. Crowdsourcing has emerged as a key practice in this transformation, enabling firms to outsource complex tasks to a distributed crowd of individuals. Among its various forms, innovation crowdsourcing platforms act as intermediaries that connect companies with skilled crowd workers to generate innovative solutions.

This transformation affects traditional HR functions, including job design, workforce planning, recruitment, training and development, performance management, compensation, and legal considerations. The transient and decentralized nature of crowd work presents important challenges for HRM scholarship, requiring renewed attention to the social and economic changes produced by digital labor markets. Building on the view that HRM encompasses the activities, policies, and practices designed to acquire, develop, and retain an effective workforce, organizations must critically assess both the opportunities and challenges associated with this collaborative online model. However, the extent to which crowdsourcing platforms support HRM functions, and the conditions under which they create value for firms, remain underexplored.

This study examines whether companies use crowdsourcing platforms as a substitute for traditional innovation outsourcing, particularly problem-solving outsourcing, or whether they seek synergies between their internal resources and those available through the platform. To explore this question, we apply Q methodology to analyze discourse surrounding Kaggle, a leading crowdsourcing platform specialized in data science, and its impact on key HR areas such as job design, workforce planning, recruitment, training and development, performance evaluation, compensation, and legal considerations. The study aims to provide deeper insight into the value these platforms offer to client companies, as well as the challenges involved in collaborating with innovation crowdsourcing communities.

Emerging technologies are often evaluated both as tools used by individuals and as developments that may reshape workplaces and society. While research on technology acceptance explains adoption primarily through individual perceptions such as usefulness or ease of use, studies of emerging technologies highlight broader ethical and social concerns. Existing research provides limited understanding of how these projected futures shape which evaluative logics become salient when individuals assess emerging technologies.

This study introduces the concept of evaluative framing to examine how the determinants of technology acceptance vary depending on whether technologies are evaluated from an individual or societal perspective. Using occupational exoskeletons as an empirical context, we conducted a survey of 150 workers who evaluated the same technology under two evaluative frames: a personal-use perspective and a societal perspective. 

The results show that perceived usefulness dominates acceptance when the technology is evaluated from an individual perspective, whereas ethical and social considerations become more influential when the technology is assessed from a societal perspective. These findings suggest that the evaluation of emerging technologies depends on the interpretive frame through which they are considered. The study contributes to research on technological diffusion by highlighting how individual adoption judgments incorporate broader societal considerations.

• Presentation of the PhD research to the stakeholders of the chair during annual scientific councils and steering committees

• Participation in open-innovation and public dissemination events (e.g., Hub Innov 2025)

• Member of working group 2 "Institutional level - Wellbeing Practices in Research Institutions"

• Assistance with the organization of ReMO Ambassador Training Schools

• Participation in the 1st ReMO Conference in Budapest, Hungary, 2023

• Responsible for deliverables on cost-benefit analysis and ethics

• Organization of workshops for stakeholders

• Communication of results to the general public on international business events (e.g., InnoTrans 2022, STREAM final event)