David Marsden Best Paper Prize 2026

Jury members: Uschi Backes-Gellner, Wenzel Matiaske, Héloïse Petit

Preliminary remarks

This year’s call for entries for the Best Paper Prize in memory of David Marsden was once again very encouraging. We received a total of 46 submissions across SASE networks and Mini-Conferences. As in previous years, entrants were asked to include a short cover letter explaining how their research relates to David Marsden’s work. The studies covered a wide range of topics: the labour-market consequences of automation, AI, and platform work; trade unions and worker voice; inequality and discrimination across gender, race, and credentials; precarious work in the Global South; comparative corporate governance and Islamic finance; global value chains; and cross-border labour standards. The methodological spectrum was equally broad, from qualitative case studies and field experiments to econometric analyses of administrative and panel data. After preselecting submissions with attention to thematic focus and relevance to the call, 29 papers were carefully reviewed by the jury. We would like to particularly acknowledge Martin Schneider’s work in the first step, who was again a great help in compiling the shortlist.

All jury members read the 29 shortlisted papers, exchanged their assessments, and consolidated the results in three online meetings. Among all the shortlisted papers, one submission stood out as a clear candidate for the prize. This is Hoyeon Lee’s paper on Polarization and Dualization in Korea. In the jury’s view, the paper is well connected to David Marsden’s intellectual tradition and addresses a substantive contemporary question with both theoretical ambition and empirical care.

The prize

This year’s prize is awarded to Hoyeon Lee (New School for Social Research) for his paper “From Polarization to Dualization: How Stratification Filters Automation Shocks in Korea.” The paper takes up one of the most consequential debates in contemporary labour-market research—the distributive consequences of automation—and asks why the canonical predictions of Routine-Biased Technological Change (RBTC) do not materialise in the most robot-intensive economy in the world. South Korea operates more than 1,000 industrial robots per 10,000 manufacturing workers, more than four times the U.S. level. Standard task-based theory would predict pronounced employment polarisation and rising wage dispersion. Korea exhibits neither: middle-skill employment has been stable, the low-wage service sector has contracted, and aggregate wage inequality has compressed during the period of most intensive robot adoption. The author treats this not as a Korean anomaly but as a theoretical opportunity to identify the institutional scope conditions of task-based frameworks.

The central argument is that the labour-market consequences of automation are not governed by the technical properties of tasks but by the institutional architecture through which labour markets allocate risk, security, and reward. What gives the paper its distinctive analytical traction is that it works on two levels simultaneously. On the one hand, it characterises the institutional position of the individual worker—wage-setting power and employment protection—through a newly constructed Labor Market Stratification Index (LMSI). On the other hand, it locates the firm itself within the broader organisation of production—whether it stands as a core firm or as a peripheral supplier in subcontracting networks and value chains. In Korea’s segmented coordinated market economy, large core firms with strong internal labour markets absorb technological pressures through internal reorganisation, while adjustment costs are shifted down the supply chain to subcontracted and non-standard workers. This dual focus resonates strongly with Marsden’s insight that employment outcomes are jointly produced by firm-level organisational rules and the broader institutional environment in which firms operate. 

The empirical strategy is, in our view, particularly elegant. The paper uses worker-level panel data from the Korean Labor and Income Panel Study (KLIPS), matched with a regional shift-share measure of robot exposure derived from International Federation of Robotics data and the Korean Census of Establishments. The author thus mobilises microdata of the highest quality to address questions that have traditionally been investigated at the aggregate level. By moving from aggregate narratives down to within-individual variation over time, the analysis recovers mechanisms that aggregate statistics conceal—most strikingly, that the apparent compression of the wage distribution is partly a compositional artefact of selective exit, as workers in weaker institutional positions are pushed out of standard wage employment into precarious self-employment. The empirical work is meticulously executed: individual and year fixed effects, regional clustering, a careful decomposition of the LMSI, a non-manufacturing placebo test, regional sensitivity analyses, alternative clustering schemes, and an explicit counterfactual exercise on inequality. 

Although still a working paper in development, the submission already stands out for the theoretical ambition and the institutional reasoning that organise its empirical work, in keeping with Marsden’s own style as a comparative institutionalist and employment-systems theorist.

In sum, this paper embodies the core aims of the David Marsden Prize. It addresses a topic of high contemporary relevance at the intersection of labour markets, technology, and organisational practices; it combines a careful analysis of the individual employment relation with an analysis of the firm’s position in hierarchical production networks; it mobilises microdata of the highest quality to address macro-level theoretical questions; and it connects back meaningfully to the intellectual legacy of David Marsden by advancing our understanding of how institutional structures—at the level of the worker, the firm, and the inter-firm production system—shape the distributive consequences of technological change.

We warmly congratulate Hoyeon Lee. We thank all participants for their submissions and for the opportunity to engage with their work.