In Germany, as in other highly developed economies, the demand for skilled workers is expected to rise further in the coming years (Burstede & Tiedemann, 2025). Demographic change makes the mobilization of both domestic and international talent increasingly necessary, highlighting the importance of recruiting international students and supporting their sustainable integration into the German labour market.

English-taught degree programmes play a key role in attracting international students, as they reduce the frequently perceived barrier of mandatory German-language instruction. They are widely regarded as a central instrument of the internationalisation of German higher education, enabling universities to recruit international students and academic staff.

In recent years, both the number of English-taught programmes and their share of the overall study provision in Germany have increased significantly (Netz et al., 2025). Between 2009 and 2025, the number of English-taught bachelor’s programmes increased almost sevenfold (from 63 to 424), while the number of master’s programmes increased more than fivefold (from 356 to 1,928). At the same time, a concentration in specific disciplines and fields of study can be observed. In particular, mathematics and natural sciences, economics and law, and engineering show a high proportion of English-taught programmes.

German higher education policy not only aims to increase the recruitment of international students but also focuses on ensuring study success and facilitating the transition into the German labour market (DAAD & BMBF, 2023; HRK, 2024; BDA, BDI & HRK, 2024). Targeted and coordinated support measures are essential for a successful transition into employment. In addition to subject-specific qualification, these measures include the systematic development of German language skills, career guidance services, access to practical experience, and the reduction of structural and (migration-)legal barriers. Preparing international students for access to the German labour market and supporting their successful transition into employment constitute a societal task that is to a large extent co-shaped by higher education institutions. Universities act as key actors at the interface between study programmes, career orientation, and labour market integration.

Against this background, the German Centre for Higher Education Research and Science Studies (DZHW), commissioned by the German Academic Exchange Service (DAAD) and funded by the Federal Ministry of Research, Technology and Space (BMFTR), is conducting the study “English-Taught Degree Programs – German Labour Market” from January 2025 to July 2026. The aim of the study is to analyse the extent to which English-taught programmes contribute to mitigating skills shortages in Germany and which programme designs are particularly suitable for unlocking the potential of international students for the German labour market. The focus is on English-taught STEM programmes and their relevance for the recruitment, education, and labour market integration of international students after graduation.

To address the central research interest, three sub-questions were formulated:

1. What role do English-taught degree programmes play in attracting international students?
2. How is the development of German language skills integrated into programme structures during and after the course of study?
3. What different language models exist in comparable programmes, and what effects can be observed with regard to the labour market integration of international alumni?

The ESDA study is designed as a qualitative survey and is primarily based on the analysis of empirical interview data. For in-depth analysis, semi-structured expert interviews and group discussions are conducted with different stakeholder groups at German higher education institutions. These include representatives of university leadership, programme directors and teaching staff, employees of central university services—particularly international offices and career services—as well as alumni of English-taught programmes.

In addition, the study includes an analysis of different models of English-taught STEM programmes and their specific features with regard to measures supporting the development of German language skills among international students. On a descriptive level, the language models implemented in these programmes are examined based on information provided on university websites. This analysis aims, among other things, to identify the pathways through which students gain access to English-taught STEM programmes (e.g. secondary school qualifications, preparatory colleges, or prior studies in their home countries).

The conference contribution presents key findings from the ESDA study and derives practical recommendations for higher education institutions, higher education and education policy actors, funding organisations, as well as international students and academics.

Budgeting is a central instrument of governance and strategic steering in higher education (Gibson, 2009). Beyond allocating financial resources, it shapes relationships between universities, public authorities, and broader socio-economic actors (West, 2023). In many systems, including Morocco, traditional input-based models have prioritized compliance and administrative control (Ayad, 2025; Cruz, 2022). While ensuring financial oversight, such models often restrict structured dialogue and universities’ capacity to act as strategic partners in socio-economic development.

In response to these challenges, Morocco introduced performance budgeting through performance agreements (PAs), locally termed contractualization, in 2009 under the Emergency Program (2009–2012), a reform initiative designed to accelerate strategic priorities and address structural inefficiencies in higher education (Cour des Comptes, 2016). This reform followed the publication of a landmark report by the Supreme Council of Education (2008), which identified structural weaknesses such as limited research capacity and high graduate unemployment and called for closer alignment between national priorities and university strategies through strengthened partnerships between public universities and the Ministry of Higher Education, Scientific Research and Innovation (MHESRI; Conseil Supérieur de l’Enseignement, 2008).

Despite ambitious partnership goals, empirical evidence on the effectiveness of PAs in fostering engagement remains limited, and systematic research on their practical implementation is scarce (Cour des Comptes, 2016). Therefore, this study addresses this gap by analyzing the design, preparation, negotiation, and implementation of PAs across three policy phases between 2009 and 2025. It examines (1) how state–university relationships are structured through these agreements, and (2) the conditions under which PAs foster engagement and collaborative state-university partnership.

To provide policy context, PAs were introduced in 2009 as a governance instrument to strengthen coordination and strategic alignment between the Ministry and public universities. Under this framework, the Ministry defines national priorities, and universities submit project-based proposals reflecting their institutional profiles and capacities while aligning with the priorities set by the Ministry. These proposals are negotiated and formalized through contracts in which the state commits financial resources and universities commit to achieving agreed objectives. PAs thus institutionalize a shift from hierarchical steering toward negotiated coordination in state–university relations. In this study, state–university partnership is defined as two-way collaborations grounded in mutual respect for distinct roles and capacities, while engagement refers to universities’ active, mission-driven interactions within such partnership (Gutierrez et al., 2025).

PAs were subsequently mobilized to operationalize Framework Law 51-17 (2021) and the National Plan for Accelerating the Transformation of Higher Education, Scientific Research, and Innovation (2022–2025). These successive reform phases reflect an evolving attempt to move beyond uniform, top-down governance toward more dialogical arrangements in which universities actively contribute to national development priorities.

Conceptually, the study approaches PAs not as static policy documents but as dynamic budgeting and governance processes. It adopts a hybrid analytical framework combining process evaluation (Moore et al., 2015) and strategic contingency theory (SCT; Crozier & Friedberg, 1980). Process evaluation provides a structured lens to examine how PAs are designed, negotiated, and implemented across reform phases (2009–2025), while SCT illuminates the underlying dynamics by focusing on power relations and strategic behavior. Together, this framework conceptualizes universities as local orders in which governance arrangements are continuously interpreted and adapted in practice.

This study adopts a multi-phase qualitative design based on purposive sampling (Creswell, 2013). Data were collected at two levels. At the institutional level, four public universities implementing PAs across the three policy phases were selected to ensure geographic diversity. At the actor level, semi-structured interviews were conducted with seven policymakers from the MHESRI and with 24 university leaders and administrators in four universities involved in the negotiation and implementation of PAs across the three phases.

While the data remain under analysis, preliminary findings suggest that PAs function as structured platforms for formal dialogue between the Ministry and public universities. However, their capacity to foster deeper engagement depends on three interrelated conditions: (1) committed leadership, supported by sustained political and technical backing and clear communication; (2) purposeful policy design, grounded in contextual awareness and oriented toward substantive outcomes rather than procedural compliance; and (3) adaptive implementation, incorporating continuous monitoring, and organizational learning throughout the policy cycle.

The rapid diffusion of artificial intelligence is reshaping both national higher education systems and the internal organization of universities. Governments increasingly frame AI as a strategic domain connected to economic competitiveness, geopolitical positioning, and regulatory sovereignty. At the same time, universities are expected to supply research, talent, and ethical guidance while adapting their own governance, teaching, and operational practices to AI-enabled tools. Despite this growing interdependence, research has typically treated states and universities separately: comparative policy studies emphasize national strategies, while higher education scholarship focuses on institutional change without systematically linking it to macro-level governance regimes.

This separation is notable given longstanding evidence that universities are deeply shaped by public policy through funding, regulation, and governance instruments (Ansell, 2008; Capano & Pritoni, 2020; Döhler et al., 2025), and that state traditions influence how higher education systems evolve (Malinovskiy & Shibanova, 2023; Andres et al., 2024). At the same time, universities -especially research-intensive institutions - actively contribute to policymaking through expertise, elite formation, and agenda setting (Marginson et al., 2023). AI therefore represents a domain where this bidirectional relationship becomes particularly visible.

Building on comparative AI policy research, we conceptualize national strategies along three dominant orientations according to Guenduez, Fuchs & Demircioglu (2025): (1) exploiting AI’s innovation and competitiveness potential, (2) capacity building through skills, education, and research infrastructure, and (3) governance and regulation aimed at ethical and trustworthy AI. These orientations coexist but vary in relative emphasis across countries. They are also embedded in broader political dynamics described as both a “space race” for technological leadership and a “race in regulation” over standards and control (Ulnicane et al., 2021; Smuha, 2021).

Universities translate these policy frameworks into organizational responses. Existing studies show that institutions differ markedly in their approaches to GenAI adoption, ranging from restrictive integrity-focused policies to proactive curricular integration and partnership models (McDonald et al., 2025; Frumin, Terentev et al., forthcoming). Moreover, universities increasingly depend on private-sector platforms and data infrastructures, reshaping governance and autonomy (Komljenovic, Sellar & Birch, 2025).

We therefore conceptualize universities as co-shapers - not merely recipients - of AI governance. Their public communication provides observable traces of strategic positioning and priority setting. The central research question is: How are universities’ GenAI agendas aligned with national AI policy frameworks across different governance contexts? Sub-questions examine what themes dominate university communication, whether these themes cluster according to governance regimes, and how institutions reproduce or adapt state priorities.

The study aims to contribute theoretically by proposing a co-evolutionary model of state=university relations in technological governance; empirically by mapping cross-national variation in institutional GenAI agendas; and methodologically by demonstrating how computational text analysis can be combined with comparative policy typologies to study higher education systems at scale.

First, we constructed a corpus of university communications using automated web scraping of English-language content from approximately 2,400 institutions listed in THE global ranking. Keyword-based searches targeted references to generative AI and related technologies. Second, we manually coded training set established category definitions, and large language model-assisted classification was used to scale annotation. We focused on “news” and press-style texts.Third, we linked universities to national AI governance contexts. Countries were grouped through hierarchical clustering based on comparative policy indicators developed by Guenduez et al. (2025). This strategy allows comparison across governance regimes rather than single-country cases, following recommendations in comparative higher education governance research (Capano & Pritoni, 2020). Fourth, we applied a Structural Topic Model (Roberts, Stewart & Tingley, 2019). STM extends classical topic modeling (e.g., LDA) by incorporating document-level covariates - in our case, country cluster membership.

Prelininary results are the following. University GenAI communication is dominated by research and innovation narratives, emphasizing projects, laboratories, and partnerships. This supports the view that universities continue to frame AI primarily through their research mission, regardless of governance context. Themes related to regulation, ethics, academic integrity, and technological risks are present across all clusters, reflecting shared sector-wide concerns similar to those documented in studies of institutional GenAI policies (McDonald et al., 2025). Stronger alignment appears in capacity-building regimes, where universities more frequently emphasize training programs, curriculum redesign, and skills development consistent with national policy priorities (Guenduez et al., 2025).