<

Small universities are overturning big research – an interview about Romanian academic networks with Adrian Hatos

Adrian Hațos - interviu despre universitățile mici în cercetare

After finishing my conversation with Adrian Hatos, I sent at least five messages to arrange for some coffee. And I had some good reasons. Sociologist and former president of the Science, Innovation and Technology Committee of the Romanian Senate, Adrian Hatos studies collaboration in the Romanian academic system through the lens of network analysis. I talked to him about a study that overturned some of my ideas about what I thought academic influence meant in Romania. It is not just money or historical prestige that matters, but also the position in a network of relationships that Adrian Hatos and Smaranda Liana Cioban-Kudelca made visible in their book „"Centrality and influence in research in Romania: What does network analysis say about the participation and success of research entities in the "Centers of Excellence" competition?"«.

„"Regardless of the calculation method, four institutions are at the core, the most influential in this whole story," says Prof. Dr. Adrian Hatos. Four universities form the core of Romanian research participating in the national "Centers of Excellence" competition in 2024: Bucharest Academy of Economic Studies, University of Bucharest, Technical University of Cluj-Napoca and Transilvania University of Brasov.

 

 

The two researchers analyzed the structure and dynamics of scientific collaboration networks, measuring how connected institutions are to each other and what are the factors that determine success in obtaining funding. But there are some major surprises in this ranking: does at least one name surprise you? Me too! But what about a few absences?

Suceava and Transylvania: "It is the result of management"„

"Transilvania University Brasov is at the core of the network by all algorithms – a medium-sized university without huge funding, but with diversified participation in several fields. So is University of Suceava. "The two are punching well above their weight, being three times smaller than other higher-rated institutions." Adrian Hatos also mentions University of Craiova, which ranks fourth in national influence and leads in brokerage, the ability to connect others with each other.

How do we explain this? Adrian Hatos has a theory: "These are comprehensive universities, they have medical specialties and engineering specialties, basically they are universities that also have engineering faculties, so they participate in various, much more diverse partnerships. You also enter with engineers, you also enter with doctors." While the large traditional universities, Babes-Bolyai for example, "I'm going to cover a few main areas.".

But the stronger argument may be another: "In Suceava it is clearly the result of a management that manages to produce more intense participation," says Hatos. "I don't think it's luck. I think it's also a specific policy of some universities to promote, to stimulate quality research and research partnerships."„

See also: The place of universities in the Universitaria Consortium in academic networks

University of Bucharest (UB)

2nd place nationally in overall influence: Eigenvector Score: 0.94 (close to the maximum of 1.00)

  • 80 direct collaborations: 2nd place in Degree Centrality)
  • Part of the hard core: One of the 4 central institutions of Romanian research
  • Full partnership: Collaborate with UBB in all 6 areas of the competition
  • Strong Brokerage: 2nd Place – connects many different groups

"Alexandru Ioan Cuza" University of Iași (UAIC)

The most versatile university in Romania: Multiplexity Score: 3.13 (1st place nationally)

  • Multi-super-bridge: Connects the network in three ways – centrally within its own group (hub), outwardly (border), and between different groups (betweenness)
  • Part of the general core: central influence in the network
  • Diversified partnerships: Links in 5-6 fields with UBB, UVT and UCV

West University of Timisoara (UVT)

Second most diverse: Multiplexity Score: 2.96 (2nd place nationally)

  • Multi-bridge: Connects the network in three ways – hub (central in the cluster), border (connects outwards) and betweenness (connects between different sub-communities)
  • In the extended nucleus: part of the central structure
  • Collaborations in 5 areas with UAIC, UB and UBB

University of Craiova (UCV)

Absolute leader in brokerage: 1st place nationally!

  • Best Deck: Deck Score 832.87 (even beats UB!)
  • Highest betweenness potential: Betweenness Centrality 1593.74 (1st place)
  • Good for indirect connection: Proximity 0.61 (3rd place, tied with UBB)
  • 72 direct collaborations (3rd place nationally)
  • High influence: Eigenvector 0.85 (4th place)
  • In the general core: confirmed influence in the central structure

Babeș-Bolyai University of Cluj-Napoca (UBB)

Stable and focused player in all areas

  • Full partnership: Collaborates with UB and UAIC in all 6 areas
  • 70 direct collaborations (according to Weighted Degree Centrality = 147)
  • Strategic bridge: 35 internal + 35 external links (dual hub-border role)
  • In the extended nucleus: part of the central structure

Bucharest Academy of Economic Studies (ASE)

Part of the "big four" — the core of research

  • One of the 4 central institutions (alongside UB, UNITBV, UTCN)
  • Brokerage potential: 5th in effective size, meaning the number of unique connections an institution has
  • Border bridge: Connects different groups on its periphery
  • Good diversity: 6th place in multiplexity

"Lower Danube" University of Galati (UDJG)

Present in the general nucleus — confirmed by Discrete Ucinet algorithm (a conservative method that strictly separates the core and the periphery, including only the most central actors)

"Lucian Blaga" University of Sibiu (ULBS)

In the extended core — one of the 13 institutions in the extended core, according to the Coreness score.

"Ovidius" University of Constanta (UOC)

Strategic border bridge

  • In the Top 10 for Betweenness: 7th place in Betweenness Centrality
  • It has a good role in connecting different groups
  • Brokerage potential

Analysis performed with NotebookLM, Kimi K2.5 and verified by me.

How Romanian research was mapped

Assistant Professor Smaranda-Liana Cioban-Kudelca

The study started from a seemingly simple question: what does collaboration in Romanian research look like when you look at it as a map? Hatos and her colleague analyzed 109 projects submitted within the "Centers of Excellence" competition, a program with a budget of approximately 1.6 billion lei (over 320 million euros), carried out in six different fields, from Health and Digitalization to Food, Bioeconomy and Civil Security.

Prof. dr. habil. Adrian Hatos and ass. univ. Smaranda Liana Cioban-Kudelca looked at the competition through a lens borrowed from network science. Each participating institution is a node, and if two institutions participate together in a project, a connection is created. The two studied how connected the nodes are to each other, which ones are at the core of the research, which nodes collaborate most intensely on multiple dimensions, which nodes bridge different parts of the network that would otherwise not collaborate with each other.

Terminology guide

Network analysis uses mathematical indicators to measure how each university positions itself in the academic ecosystem. Here is what the main terms mean, explained simply:

Eigenvector CentralityHow well connected are your friends? It's not just who you're directly connected to, but how well connected your partners are. It's like in real life: being friends with someone influential increases your influence as well.

Degree CentralityHow many friends do you have? It measures the number of direct collaborations a university has. The more direct links, the greater the potential for communication and cooperation.

Closeness CentralityHow quickly does the news reach you? It indicates how close an institution is to all others in the network, including indirect links. A high value means that information flows quickly through that university.

Betweenness CentralityHow often do you go through the middle? It measures how often a university is on the „shortest path” between two other institutions. High scorers „bridge” groups that would otherwise not connect.

Bridge ScoreAre you a bridge between worlds? Identify universities that connect different sub-communities. They can be "bridge-hub" (connect a lot within their group) or "bridge-border" (connect different groups).

Brokerage / Structural HolesAre you filling the gap between them? When two institutions do not communicate directly, but do so through you, you occupy a „structural gap.” Actors in these positions are „network entrepreneurs” who can control the flow of information and opportunities.

MultiplexityHow diverse is your collaboration? It measures how many different fields a university's partnerships cover. A high value means collaborations across multiple scientific areas — from health to engineering to digitalization.

The Hard Core (Core)Are you in the center of power? The network has a core-periphery structure: a dense, well-connected core and a less integrated periphery. Being part of the core means maximum influence and priority access to resources.

Analysis performed with NotebookLM, Kimi K2.5 and verified by me.

Several important actors take the cream

Not all of the study’s findings are surprising. First, the Romanian research ecosystem is centralized and hierarchical. Second, it is controlled by large public universities, which coordinate 62% of project proposals. National research institutes participate, but rarely lead. „We have a clear distinction between organizations that dominate the research landscape and those that remain on the periphery,” explains Adrian Hatos. „Romania’s situation is one of very strong centralization: a few nodes take the cream of all the relationships, and the rest are connected to each other through these central nodes. If you want to participate, you have to connect to an important one.” Public universities and national institutes are the dominant actors, while NGOs, private companies and museums remain marginal. No private university, NGO, SRL or museum holds the role of coordinator in this competition.

At the national level, according to the Eigenvector Centrality indicator that measures the quality of connections (not just who you are connected to, but how well connected your friends are as well), Bucharest Polytechnic is the absolute leader with the maximum score of 1.00, followed by University of Bucharest (0.94) and Technical University of Cluj-Napoca (0.85). The surprise comes from the University of Craiova, placed in fourth place with the same score of 0.85, which becomes the leader in the brokerage chapter. Babeș-Bolyai University occupies fifth place (0.84), being a stable central actor in all fields.

 

 

At the same budget, connections offer an advantage

What the two authors found is that, for the same institutional size measured by budget, centrality makes the difference. In other words, between two institutions that are equally well funded, the one that is better connected will win. Statistical regressions show that universities should have been more successful if size were the only factor, but centrality mediates the impact of size. Money does not win funding.

Money helps you build connections, but ultimately it's the connections that matter when funding partnerships. Oradea and Brasov They're about the same size, the same faculties, similar budgets. One is in the core, the other is not.

Institutions with high betweenness centrality, that is, those that bridge otherwise separate groups, are 17 times more likely to coordinate a funded project than isolated ones. „The most effective strategy for increasing the chances of success is intense and diverse participation in consortium networks,” Hatos emphasizes.

Bucharest concentrates resources, but it is not Silicon Valley

„"52% of the participants are from Bucharest-Ilfov. With this concentration of resources, we should have a Romanian Silicon Valley, but it doesn't. Something is wrong there." Adrian Hatos points out that half of the national research happens within a thirty-kilometer circle around the Palace of Parliament, but without synergy. In contrast, the universities in the province that manage to break the isolation, Braşov, Craiova, Suceava, Constant, are the ones who understood that in a network it matters to be connected with everyone, not just those like you.

And even though they dominate the research landscape, universities are missing something essential. "Universities should be the place where various partners meet - INCDs, companies, hospitals. They should organize conferences, living labs, be a creative effervescence. Apart from courses and laboratories, practically nothing is done."„

Projects don't start with a contract, but with a coffee

Although the discussion was about advanced mathematical analysis, Adrian Hatos was careful to emphasize from the beginning that what the graphs and coefficients show are not abstract entities, but people. "Institutions do not always collaborate with each other. Behind institutional partnerships are human relationships. These partnerships are not formed randomly, but are the result of affinities, a history of collaboration."„

Network analysis shows us the map. But maps don't tell us how the roads were built, for that we have to remember the people and the coffees.

CategoriesUncategorised

Leave a Reply

Your email address will not be published. Required fields are marked *