Redeﬁning Data Governance: Insights from the French University

System

Guy Melançon

1 a

, Nathalie Pinède

2,3 b

and Ugo Verdi

1,3 c

Univ. Bordeaux, CNRS, Bordeaux INP, LaBRI, UMR 5800, F-33400 Talence, France

Univ. Bordeaux, CNRS, Bordeaux INP, IMS, UMR 5218, F-33400 Talence, France

MICA - Médiation, Information, Communication, Art, MSHA – 10 Esplanade des Antilles, 33607 Pessac, France

Keywords:

Data Governance, French Universities, Collegiality, Subsidiarity Principle, Inter-Organizational

Collaboration, Multi-Tiered Governance.

Abstract:

This position paper highlights the distinctive features of French universities that render data governance within

these institutions a particularly challenging endeavor. These universities inherently operate in an exception-

ally open milieu, necessitating the conceptualization of governance as a dynamic and adaptable framework

that converges seamlessly with the governance structures of other institutions. The principle of collegiality

further mandates a distributed approach to data governance, encompassing responsibilities, rules, and proce-

dures across various levels of management. Moreover, it is essential to reevaluate the conventional viewpoint

that segregates administrative tasks from research and teaching functions. Our ﬁndings underscore the ne-

cessity for developing and executing a dynamic, multi-tiered data governance model that integrates the three

fundamental missions of universities. Given the intrinsic nature of French universities, it is imperative to

envisage governance as an evolving ecosystem of agents who assume complementary responsibilities in a har-

monized manner.

1 INTRODUCTION

In the current era, characterized by a data-intensive

landscape (Bouchez, 2014) and an innovation-driven

knowledge economy (Wessels et al., 2017), universi-

ties emerge as pivotal contributors to research. Their

role is further accentuated by the imperative of sus-

tainability (Musselin, 2006). Like all organizations,

they must anchor their strategic and managerial de-

cisions in the analysis of data that encapsulates their

operations. This goal, propelled by the generation and

utilization of data, necessitates the development of ro-

bust data governance within these institutions.

While some Anglo-Saxon universities have al-

ready established data governance frameworks

, their

French counterparts appear to be in a state of catch-

up, often contingent upon centralized regulatory di-

https://orcid.org/0000-0003-3193-7261

https://orcid.org/0000-0002-5381-5524

https://orcid.org/0000-0002-0562-6032

Examples include University of Wisconsin-Madison,

University of Michigan, New York University, Oxford Uni-

versity, University of San Francisco, University of Toronto

and University of Washington.

rectives. The adoption of “corporate” models of data

governance, frequently based on DMBOK guidelines

(DAMA International, 2017), might not be entirely

suitable for French universities. This mismatch un-

derscores the ongoing efforts to revamp their manage-

rial structures (Mignot-Gérard et al., 2023) and high-

lights the need for a tailored approach to data gover-

nance in this speciﬁc milieu.

A deeper understanding of the context and oper-

ations within academic settings will lay the ground-

work for developing a dual-purposed methodology:

ﬁrstly, to establish data governance, detailing its ar-

chitecture and operations, and secondly, to assess it by

evaluating the maturity level of the institution’s data

governance practices.

This position paper delves into the unique charac-

teristics of French universities, positioning data gov-

ernance as a signiﬁcant and distinct case study. Our

analysis is grounded in the experience of the Uni-

versity of Bordeaux, which is currently implement-

ing data governance strategies. We pay particular at-

tention to the ANR ACT initiative (ACT stands for

“Augmented university for Campus and world Tran-

sition”, the project being funded by the French Na-

Melançon, G., Pinède, N. and Verdi, U.

Redeﬁning Data Governance: Insights from the French University System.

DOI: 10.5220/0012699300003690

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 26th International Conference on Enterprise Information Systems (ICEIS 2024) - Volume 2, pages 733-738

ISBN: 978-989-758-692-7; ISSN: 2184-4992

733

tional Research Agency), a pioneering project that en-

compasses a suite of strategic institutional projects,

all dedicated to the principles and practices of data

governance.

2 THE FRENCH CONTEXT

In our examination, we critically analyze French uni-

versities as distinct entities, underscoring traits that

not only set them apart from the corporate sector but

also render the French university system unique in

its essence. Despite recent signiﬁcant transformations

within these institutions, French universities continue

to navigate the delicate balance between striving for

autonomy and adhering to the mandates of a central-

ized state (Vinokur, 2008; Musselin, 2017). This in-

tricate dynamic results in speciﬁc demands for data

governance that are unique to the French context

(Pierronnet and Blanchard, 2021). This duality of in-

dependence and state directives shapes the framework

within which these universities must operate, inﬂu-

encing their approach to managing and utilizing data.

2.1 Interconnected Governance: The

Open and Collaborative Structure

of French Universities

The organizational complexity of French universities

often astonishes those unfamiliar with the system. A

hallmark of these institutions is their collaboration

with a variety of supervisory bodies, both local (such

as Institut Polytechnique) and national (like CNRS

Inserm

, or INRAE

). While research laboratories are

physically situated within university campuses, a sub-

stantial portion of the staff utilizing these facilities,

operating local infrastructures, or managing adminis-

trative processes are not university employees. This

openness fosters a deep entanglement of personnel

and daily operational protocols from various institu-

tions, a complexity that transcends simple contractual

arrangements and is ingrained in the very fabric of the

university system.

French universities, as state entities, must harmo-

nize their strategies not only with the state’s objec-

tives but also operate within a strict regulatory frame-

work. A signiﬁcant number of the degrees they confer

CNRS is the Centre National de la Recherche Scien-

tiﬁque

Inserm is the Institut national de la santé et de la

recherche médicale

INRAE is the Institut national de recherche pour

l’agriculture, l’alimentation et l’environnement

are national diplomas. The Ministry of Higher Edu-

cation exercises oversight and approval of university

curricula to maintain nationwide uniformity. This en-

sures that the value of a diploma is consistent, irre-

spective of the issuing university.

However, France’s state-led research policy is del-

egated to speciﬁc research organizations (state oper-

ators) like the CNRS, with other bodies overseeing

research in particular domains (Inserm for health, IN-

RAE for environmental sciences, etc.). These opera-

tors’ policies signiﬁcantly inﬂuence the universities’

research strategies, necessitating alignment with na-

tional priorities. Furthermore, the state’s data policy

shapes the policies of these organizations and univer-

sities (Ministère de l’Enseignement Supérieur, de la

Recherche et de l’Innovation, 2021).

This open and interwoven organizational structure

profoundly affects the intellectual property rights of

research outcomes and the generated (research) data.

Many projects produce their data while also utilizing

external data from national databases or from institu-

tional and industrial partners. Given the diverse in-

stitutional afﬁliations of team members, data must be

regarded as a shared asset. This complexity extends

to administrative data related to project management,

hiring processes, and more.

Consequently, this mode of operation necessitates

a conceptualization of university data governance that

is intricately connected with that of organizations like

CNRS, Inserm, and partner entities. Thus, when es-

tablishing governance structures, it is crucial to con-

sider not only university stakeholders but also broader

ecosystem actors at both decision-making and orga-

nizational levels. Governance rules and procedures

must be harmonized with those of ecosystem partici-

pants, mirroring challenges akin to those posed by the

European Data Protection Regulation (GDPR). While

harmonization might sometimes be project-speciﬁc,

involving ad hoc partners, achieving consistency at an

institutional level with organizations such as CNRS or

Inserm, for instance, can be essential.

This scenario stands in stark contrast to gover-

nance models prevalent in Anglo-Saxon universities

(Jim and Chang, 2018) or those proposed in aca-

demic literature (Abraham et al., 2019), highlight-

ing the unique challenges and considerations in the

French context.

2.2 Balancing Governance: Collegiality

and Subsidiarity in University

Structures

Universities are deeply rooted in a tradition of col-

legiality, which inherently precludes a top-down,

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

734

Figure 1: A look at the French academic ecosystem. Multiple supervisory entities shape data governance through their

scientiﬁc policies and operational connections with research endeavors. This inﬂuence extends to collaborations with local

authorities or industrial partners, driven by institutional mandates or speciﬁc projects. The interconnectedness of university

projects, units, and components necessitates a nuanced, multi-tiered approach to governance, necessitating a tailored applica-

tion of the subsidiarity principle to accommodate this complex structure effectively.

strictly hierarchical governance model (Musselin,

2006). A notable challenge arising from this tradi-

tion is that, despite the coherence and intelligence

demonstrated by top-level management teams in col-

laboration with central services, directors of lower-

level units (such as teaching departments and research

units) often play a lesser role in decision-making or in

endorsing the institution’s policies (Chatelain-Ponroy

et al., 2012).

The appointment of a data steward for speciﬁc

domains or research areas is complicated by re-

searchers’ tendency to act independently, and for in-

stance freely select data storage infrastructures (for

example, recherche.data.gouv.fr at a national level, or

Zenodo at European level). This independence ex-

tends to course materials, which may be hosted either

on local platforms or national portals like France Uni-

versités Numériques (FUN).

Project leaders generally establish their data poli-

cies within their own areas, regardless of the project’s

strategic signiﬁcance to the university. Incidentally,

these projects often include staff from various orga-

nizations and might even receive joint funding from

these entities. This scenario directly affects how

upper-tier data governance bodies should tackle gov-

ernance issues to ensure alignment between overarch-

ing institutional rules and speciﬁc project-level best

practices.

In this light, the principle of subsidiarity is indis-

pensable, providing a framework for allocating com-

petences between these two governance tiers. The

challenge lies in devising a harmonious integration

that is applicable across the university while still be-

ing pertinent to individual projects.

Moreover, the principle of subsidiarity might be

the optimal strategy for incorporating the multi-site

aspect of university campuses. Local administrators

and elected ofﬁcials may resist policies that man-

date the centralization of data, favoring instead a dis-

tributed model that grants them some measure of con-

trol and inﬂuence over the data they generate or uti-

lize.

2.3 The University, The Three-Headed

Cerberus of Education

From a business standpoint, data governance is typi-

cally geared towards enhancing efﬁciency and prof-

itability. However, in the context of universities,

while efﬁciency and sustainability remain important,

there is also a strong emphasis on social goals such

as democratizing access to science and ensuring its

affordability for the public. Additionally, universities

are bound by the requirements of transparency in pub-

lic administration.

Universities’ objectives span three primary mis-

Redeﬁning Data Governance: Insights from the French University System

735

sions: research, teaching, and administration (with

universities acting as implementers of state education

policy). At ﬁrst glance, this tripartite mission might

suggest a need for a three-pronged approach to data

governance.

• Research data is notably diverse, varying in ori-

gin, use, and particularly in terms of its produc-

ers and users, who are often afﬁliated with dif-

ferent supervisory bodies and come from vari-

ous disciplinary backgrounds. This data also re-

ﬂects the practices of different research commu-

nities. Moreover, national and European open sci-

ence initiatives play a signiﬁcant role in shaping

the governance of research data at the institutional

level.

While universities might leverage insights from

the Research Data Alliance (RDA) which primar-

ily aligns with open science (Madison, 2020), the

approach here diverges, focusing on integrating

data governance within the strategic framework of

universities as socio-economic entities. Our ap-

proach critically examines the segmentation of re-

search data management as an isolated university

function.

• Teaching-related data, in contrast, is more local-

ized, originating from educational activities in-

volving university students. This data type is sub-

ject to multiple regulatory frameworks: the Eu-

ropean GDPR, due to its inclusion of personal

data (like student proﬁles and academic records),

and national regulations regarding the preserva-

tion and archiving of educational data, among oth-

ers.

• Institutional data, usually sourced from the uni-

versity’s information system, most closely resem-

bles corporate data. However, this comparison is

somewhat superﬁcial, as we will explore further

in the subsequent sections.

In conclusion, while the framework of data gover-

nance in universities may draw some parallels with

corporate models, it is distinct in its complexity

and scope. Universities must navigate a unique

landscape where data governance intersects with di-

verse academic disciplines, regulatory environments,

and social responsibilities. This intricate matrix de-

mands a bespoke approach to data governance, one

that harmoniously integrates the nuances of research,

teaching, and administrative data while aligning with

broader societal and educational objectives.

3 REDEFINING DATA

GOVERNANCE IN FRENCH

UNIVERSITIES: A

MULTIFACETED APPROACH

The insights gathered from our observations under-

score the necessity for data governance in French uni-

versities to be conceptualized as a multi-tiered frame-

work. Notably, existing literature on data governance

in primarily corporate contexts, such as the work of

(Alhassan et al., 2018) and (Abraham et al., 2019), or

in the realm of higher education as examined by (Jim

and Chang, 2018), does not adequately address the

requirement for a distributed governance model that

can effectively accommodate collegiality. The princi-

ple of subsidiarity is critical here, as it must empower

the university to manage its collegial structure, while

simultaneously ensuring that individual projects are

consistent with the university’s overarching data strat-

egy.

3.1 Strategic Project Identiﬁcation and

Advancing Data Literacy

In a governance structure envisioned as multi-tiered,

it is essential for projects to extend beyond sim-

ple data management plans to include comprehensive

managerial activities. These activities should be in

harmony with and contribute to the broader strategies

of the afﬁliated institutions. The role of data liter-

acy, as a spectrum of digital competencies, is increas-

ingly recognized in universities (Verdi and Le Deuff,

2020). This literacy should be leveraged to integrate

data competencies with strategic data governance, en-

abling project leaders to understand and align with

the university’s strategic data objectives. Governance

structures, thus, need to be adept at identifying and

focusing on projects with strategic signiﬁcance.

3.2 Dynamic Governance and

Inter-Organizational Collaboration

The convergence between a university’s data gover-

nance and that of its ecosystem partners is an integral

aspect of the inter-organizational governance dimen-

sion. Universities often excel in crafting multiparty

agreements involving diverse partners. This expertise,

encompassing strategic, legal or functional domains,

should be extended to the realm of data governance.

Data considerations should become a staple in long-

term collaborative programs, ensuring continuous di-

alogue with both institutional and industrial partners.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

736

3.3 Unifying Missions Through Data

Governance

Data governance in universities should act as a uniﬁer

of the institution’s primary missions: administration,

research, and teaching. Current trends show univer-

sities harnessing their research capabilities to address

critical issues like energy transition, sustainable de-

velopment, and social responsibility.

The ANR ACT SmartMob/Datacampus project

exempliﬁes this trend. This project, focused on map-

ping campus mobility to improve transport services.

Three organizations fund this initiative: Université de

Bordeaux, Bordeaux INP and Fondation Université

de Bordeaux. Obviously, Bordeaux Metropole shows

interest in the initiative and could possibly wish to

reuse part of the research including data, and even

share data produced by IoT devices installed on their

facilities surrounding the campuses.

It demonstrates how administrative data can be

transformed into research data and educational re-

sources. Within the scope of this project, administra-

tive data detailing user mobility on campuses is used

for research, complementing scientiﬁc publications.

This data is also customized for educational purposes,

utilized in classroom settings, term projects or student

internships. Furthermore, additional administrative

data from Université de Bordeaux and Bordeaux INP,

like course schedules is integrated into the project to

aid in developing predictive models, such as those

forecasting peak campus hours.

This multifaceted use of data underlines the need

for a uniﬁed governance approach that accommo-

dates administrative, research, and teaching objec-

tives seamlessly.

In summary, this practical example vividly illus-

trates the convergence of administration, research,

and teaching around a shared data-centric project. It

also highlights the necessity for different organiza-

tions’ governance models to synergize, facilitating the

acquisition and utilization of data for comprehensive

research programs. This approach marks a signiﬁcant

departure from traditional data governance models,

advocating for a more integrated, collaborative, and

ﬂexible framework in French universities.

4 CONCLUDING REFLECTIONS:

CHARTING THE PATH FOR

DATA GOVERNANCE IN

FRENCH UNIVERSITIES

The transformative shifts that have swept through the

French university system over the past decade neces-

sitate a more structured and managerial approach to

data governance. However, these changes have not di-

minished the deeply-rooted collegial ethos of French

universities (Mignot-Gérard et al., 2023). Conse-

quently, the implementation of data governance must

be conceptualized as a dynamic and responsive pro-

cess, one that adapts to the evolving characteristics of

universities amidst ongoing transformations.

Dynamicity should be pursued as a fundamental

objective, endowing data governance with the versa-

tility to navigate diverse contexts – economic, politi-

cal, regulatory, or systemic. This adaptability is cru-

cial for realizing the full potential of a data-driven

strategy and management approach.

Dynamicity in data governance resonates with the

concepts of data lineage and the varying needs of

different data audiences. It is often necessary to

tailor data access based on the intended data usage

and user proﬁles. While data typically originates as

a singular, well-deﬁned set representing a speciﬁc

phenomenon, it often evolves into multiple datasets

through processes such as simpliﬁcation, aggregation,

or anonymization. These derived datasets are crafted

for distinct audiences, which may not be fully iden-

tiﬁed at the data’s inception. This evolving nature of

data underscores the need for ﬂexible governance that

accommodates rules and procedures to varying data

transformations and user requirements, possibly un-

known at the origin of the data lineage.

Upon reviewing existing frameworks for data gov-

ernance (e.g., (Alhassan et al., 2018); (Abraham

et al., 2019); (Gagnon-Turcotte et al., 2021); (Seiner,

2014)), it becomes apparent that none sufﬁciently en-

capsulate the nuanced characteristics imperative for a

dynamic, multi-level governance model. The intrin-

sic complexity of French universities necessitates en-

visioning governance as a ﬂuid system, where agents

assume complementary responsibilities in a harmo-

nized fashion. Yet, within this distributed system, a

regulatory framework dictates oversight by a higher

authority, ensuring coherence and alignment, while

still allowing for innovation and adaptation from var-

ious levels and peripheral players.

We are currently conducting interviews with var-

ious stakeholders, including project leaders and data-

related service managers (like those in IT and Legal

Redeﬁning Data Governance: Insights from the French University System

737

Departments). These interviews serve a dual purpose.

Firstly, they contribute to the development of a gov-

ernance maturity model speciﬁcally tailored for the

French academic context. This model is partly based

on insights gained from these interviews. Secondly,

the interviews are instrumental in identifying effective

mechanisms for establishing a dynamic, multi-level

governance architecture, a crucial aspect for evolving

academic environments.

In the realm of governance, a pivotal focus should

be placed on data literacy. This concept extends be-

yond mere skill acquisition. It encompasses a broader

understanding and contextual application of data. It

involves cultivating a mindset that recognizes the

strategic value of data, encouraging individuals to

think critically about how data can be utilized effec-

tively within their speciﬁc roles and projects. This

aspect of data literacy is about developing a deeper,

more nuanced appreciation of data’s role in decision-

making, problem-solving, and innovation. It’s about

empowering individuals to not just use data tools and

techniques, but to understand the implications of data

in the broader context of their work, the organiza-

tion’s goals, and even societal impacts. This holistic

approach to data literacy facilitates a culture where

data is not just a tool, but a fundamental component

of strategic thinking and planning.

ACKNOWLEDGEMENTS

This work was supported by a French govern-

ment grant managed by the Agence Nationale de

la Recherche (ANR) under the “Investissements

d’avenir program”, reference ANR-20-IDES-0001.

REFERENCES

Abraham, R., Schneider, J., and vom Brocke, J. (2019).

Data governance: A conceptual framework, structured

review, and research agenda. International Journal of

Information Management, 49:424–438.

Alhassan, I., Sammon, D., and Daly, M. (2018). Data gover-

nance activities: a comparison between scientiﬁc and

practice-oriented literature. Journal of Enterprise In-

formation Management, 31(2):300–316.

Bouchez, J.-P. (2014). Autour de “l’économie du savoir”:

Ses composantes, ses dynamiques et ses enjeux.

Savoirs, 34(1):945.

Chatelain-Ponroy, S., Mignot Gerard, S., Musselin, C., and

Sponem, S. (2012). La gouvernance des universités

françaises. Research report, HEC Montréal ; Cen-

tre de sociologie des organisations ; Conservatoire na-

tional des arts et métiers ; Université Paris-Est Créteil.

Gagnon-Turcotte, S., Sculthorp, M., and Coutts, S. (2021).

Les partenariats de données numériques: mettre les

bases d’une gouvernance de données collaboratives

dans l’intérêt du public. Technical report, Laboratoire

d’innovation urbaine de Montréal.

Jim, C. K. and Chang, H. (2018). The current state of data

governance in higher education. Proceedings of the

Association for Information Science and Technology,

55(1):198–206.

Madison, M. J. (2020). Data governance and the emerging

university. In In Research Handbook on Intellectual

Property and Technology Transfer, pages 364–390.

Edward Elgar Publishing.

Mignot-Gérard, S., Chatelain-Ponroy, S., Sponem, S., and

Musselin, C. (2023). Kaleidoscopic collegiality and

the use of performance research metrics. the case of

french universities. Higher Education, 85:887–918.

Ministère de l’Enseignement Supérieur, de la Recherche et

de l’Innovation (2021). Politique des données, des

algorithmes et des codes sources : Feuille de route

2021-2024. Technical report, MESRI.

Musselin, C. (2006). Are universities speciﬁc organisa-

tions? In Krücken, G., Kosmützky, A., and Torka,

M., editors, Towards a Multiversity? Universities be-

tween Global Trends and National Traditions, pages

63–84. Bielefeld: transcript Verlag.

Musselin, C. (2017). La grande course des universités.

Presses de Sciences Po.

Pierronnet, R. and Blanchard, A. (2021). La donnée : un

enjeu managérial à saisir pour les universités. In Prox-

imités et territoires aux déﬁs du management public.

Association Internationale de Recherche en Manage-

ment Public (AIRMAP).

Seiner, R. S. (2014). Non-Invasive Data Governance: The

Path of Least Resistance and Greatest Success. Tech-

nics Publications.

Verdi, U. and Le Deuff, O. (2020). La data literacy

distribuée: Périmètres déﬁnitionnels, origines docu-

mentaires, perspectives réticulaires. Les Cahiers du

numérique, 16:137 – 173.

Vinokur, A. (2008). Vous avez dit “autonomie” ? Mouve-

ments, 55-56:72 – 81.

Wessels, B., Finn, R., Wadhwa, K., and Sveinsdottir, T.

(2017). Open Data and the Knowledge Society. Am-

sterdam University Press.

ICEIS 2024 - 26th International Conference on Enterprise Information Systems

738