Engaged research and teaching respond to urgency and needs in different societal sectors
Mirko Schäfer is founder and Faculty of Science Lead of the Utrecht Data School, where many research projects are carried out using digital methods and data analysis. The research focuses on responsible data practices and AI, and on the analysis of public debates on social media. Most projects are funded by third-party funding streams and are carried out in cooperation with external partners. Hanne ten Berge (Educational Development & Training, Utrecht University) interviews Mirko about how he manages to collaborate with the field, why he does that and what it brings him.
You finance research mainly with third-party funding. So, companies and various organisations finance your research. Why do you choose this way of working?
‘I am very unsuccessful when it comes to second stream funding. I did not win a Veni. And in the Netherlands that often means a subscription to NWO rejections. So, I started to look for other possibilities of funding research. I was intrigued by the challenges companies, government organisations or NGO’s experienced with digitization and using big data or AI. What are their problems and on what questions do they need answers? It turned out that there was a large overlap with our research agenda at UDS.’
What does that mean for your research?
‘Through working with third-party funding streams, our own research agenda has changed considerably. Research now develops very differently, namely through including an understanding of the different societal sectors. For me as a researcher, it means that I have distanced myself somewhat from my “hobby”. I no longer determine my own research agenda, but am partly led by the urgent questions in society. It really is a different perspective on research. But certainly not less interesting!’
‘This way of working requires you to question your own research interests creatively and critically. You have to ask yourself ‘to what extent is what I am researching also urgent for society?’ My research - which in a nutshell is about how our society talks about technology - has become broader and at the same time more concrete through this approach. With UDS, I now often focus on how public administration deals with data and AI. As a result, our research has also become much more relevant for society, which has also increased its impact.'
How do you know what external partners need? How do you find out?
‘By going into the field and listening to the interests, questions and needs in organisations. Then it works according to the 'land and expand' principle. You start with a small project, gain trust and then you expand the offer to other areas of that organisation or sector. The small projects we usually start with are not immediately relevant to our research agenda. However, they often provide good learning opportunities for students. And more importantly, they allow us to see that an organisation has other questions to which we have an answer from the Utrecht Data School. This is how we build a relationship of trust. That works much better than trying to sell your own research. Then you are a kind of salesperson with a product that often only minimally matches the questions that are present in the societal sector.’
You say the knowledge transfer is more effective. Can you explain how you have an impact through the Utrecht Data School?
‘We at the Utrecht Data School have defined a number of principles for our research:
- We must be there where it happens. So where the change is taking place. If you want to study the social impact of Artificial Intelligence (AI) and you do it exclusively from your university desk without encountering any algorithms or people who are affected by it or any organisation that is changed by an algorithmic system, you are doing the wrong thing.
- Education and research are structurally linked. We take students to companies and organisations and they actively participate in our research. As a result, our research is mainly applied, but ultimately it leads us to questions for fundamental research.
- Our work must have impact. What we do at the Utrecht Data School must produce a visible result in the field. And I don't just mean a scientific publication, but a real change in the organisations we work for. And it also brings changes for our students. By training them in this practice, we hope to offer them a good starting point to a fantastic career.
The great thing is, the people who worked with us, but also our alumni, now work in the sectors where we do research. They take the ideas of the Utrecht Data School and the expertise with them into the field. They know our results and then implement them into their own practice.’
I am very unsuccessful when it comes to second stream funding. I did not win a Veni. And in the Netherlands that often means a subscription to NWO rejections. So, I started to look for other possibilities of funding research.
That means you have societal impact, not only through research, but also through education?
‘Exactly. And that is also how we look at education for professionals. We have an immediate impact through the professionals we train. An employee who has just graduated often has limited agency and often cannot change much in an organisation yet. So, we don't know how well our perspective on responsible AI and data science will hold up. The advantage of teaching experienced professionals is that they already have more capacity to act and can therefore better implement our results.’
‘The municipal audit office in Rotterdam, for example: In a recent report the organisation advised to use the tools of the Utrecht Data School when public management organisations start working with algorithms. Municipalities and ministries also ask us to help them to think about data issues, and we train people in the use of our tools. This enables us to see how our tools are used in practice and how policy is adapted on the basis of our research results. And we also see examples of which good data projects are actually being developed in the Netherlands and which bad ones are being terminated.’
The Utrecht Data School offers courses, programmes and trainings that have impact in the field. But you do more than that. The Utrecht Data School also develops products, such as the Digital Advisor?
'De Digitale Raadgever (Digital Advisor) is a product of the Dataworkplace, an interdisciplinary research project of the chair for public innovation (Professor Albert Meijer) and UDS together with various municipalities and provinces. The ‘Digitale Raadgever’ is a kind of cheat sheet for council members to develop knowledge about data and digitalisation projects. It actually came about by chance. In discussions with councillors, our researchers observed that elected representatives do not experience big data and AI as political issues, and especially not as a local problem. But they are. So, we wanted to help with a very accessible tool that helps asking critical questions. The municipality for which we wanted to produce it, initially had doubts about it. We were convinced that this would be a good product that would really help them. We then went on to test the tool in other municipalities and used the results to go back to them. That is what we call ‘entrepreneurial thinking’. Sometimes you have to make more effort to convince the partner, and invest in developing something which will pay off later. That requires different skills. In my opinion, academics are still far too undervalued in this respect.’
‘The ‘Digitale Raadgever’ is also a concrete example of how the work of an academic has changed. It actually is a very educational tool, but it is not a course or a lecture. We hadn't thought of it that way beforehand. For us, it was initially 'just' a means for improving deliberative processes. Yet, in this way, you do communicate the results of your own research. More and more academics are sharing their research results in this way. And that raises the question of whether, in the long term, the description of what constitutes academic work should be expanded. Doesn't it include much more than writing scientific publications and educating students?’
How do you fund products like the ‘Digitale Raadgever’?
‘The ‘Digitale Raadgever’ is one of several practical products that we developed in the Dataworkplace with and for our external partners. The Dataworkplace arose from a research application that was rejected by science funding organisation NWO. We did not receive that grant, but the external partners who participated in this application found this project so relevant that they wanted to pay for it themselves, without the NWO funds. This developed into what we call transdisciplinary research; research together with people outside science who, in our case, financed these activities.’
In your research, it is not always clear beforehand what it will yield and what an external partner will gain from it. It is not a guarantee that working with you will always produce a concrete and useful output. Isn't that a problem?
‘For some organisations - but not all - it is not a problem. But these are often organisations that have the financial resources to say this is relevant and this is what we are going to focus on. The uncertainty of not knowing what will come out of a collaboration does require policymakers or administrators to be able to think abstractly and to be able to look far into the future. The smaller an organisation, the less budget there is for this kind of uncertain projects and the more difficult it is to set up such cooperation. Small municipalities, for example, often have no time or money for research at all, they just try to do their work with the resources they have. That is why it sometimes is quite difficult for us to get assignments from them. I do believe that administrators sometimes need to have the courage to invest in these analytical and reflective processes, because it is precisely what will help them to respond to the challenges of the digital transition, for example.’
Make sure that you build up trust by really listening to what is going on at the external partners. You have to learn to speak their language. And you will sometimes have to adopt a 'humble' attitude, because what you think you are offering may be much less relevant to these organisations than you think. And you need to be patient, because this process takes time.
Is this socially engaged and externally financed research an alternative?
'No, it is only an addition to the way we work as universities, though - in my opinion - a relevant and effective one. But so-called third revenue funding cannot possibly replace traditional funding, and it is not a compensation for budget cuts we have seen over the past years. Especially the fundamental research needs funding from first and second revenue streams.'
What are challenges in working the way you do?
'While, we are generally very positive about the opprotunities our modus operandi provides, there are some challenges indeed. Third revenue stream projects need effective support from the RSO and legal deparments. The administrative efforts should not be underestimated. Setting up contracts, billing, putting together interdisciplinary teams across different faculties and departments, and to do so in time before a project is about to start is quite a challenge. Developing and keeping up the network of external partners is an ongoing effort. We invest quite some time in these activities, which are basically not recognized as part of an academic job description. It is invisible labour. Another item on this list might be the unproductive and in our opinion dated separation of support and research staff.'
What can you recommend your colleagues who also think this is an interesting way of working?
‘First identify who might be potential partners for your area of expertise and research agenda. From there on, define a problem statement and then go back to the drawing board and think about what you can contribute to the questions that are important in the field. But start small. You start with a small project and learn from it. What went well and what can we do better? I always recommend to develop a tool, product or service of which you know there is a demand for in the field, because then you have something to offer immediately instead of having to ask for money without something tangible in return. An additional advantage is that you come in as an expert.’
‘Make sure that you build up trust by really listening to what is going on at the external partners. You have to learn to speak their language. And you will sometimes have to adopt a 'humble' attitude, because what you think you are offering may be much less relevant to these organisations than you think. And you need to be patient, because this process takes time. For example, I drink many cups of coffee with different people before I actually get an assignment. But it is important that you do so. It makes you visible. I am available to people, for example by giving lectures. I see that as acquisition costs, but also as part of my vocational task to share knowledge. The same goes for journalists. They know they can call me and I will provide them with extensive information on data-related topics. It is a kind of tutoring I provide. I don't always get the credits for it in the media, but it does help building a relationship. The same applies to municipal councils, provincial executives and ministries. By now they know where to find us and they also know that we don't send invoices straight away.’
‘Another way to get in touch with the field and to be visible is our presence at trade fairs. For example, ‘Overheid 360⁰’ a congress and trade fair for data-driven public management. We, from the UU, are present with a stand that is offered to us by the organisation. We are there as a knowledge partner among the commercial exhibitors. A trade fair like that is the perfect way to speak to the field directly.’
‘And finally, I can also recommend publishing in the magazines that people in the field read. So, in the journals for professionals. Find out which magazines are relevant. For example, I give interviews to the magazines Vereniging van Vlaamse Steden en Gemeenten, De Ingenieur and others. There are also articles about our research in VNG magazine, the magazine of the Association of Dutch Municipalities, iBestuur, Binnenlands Bestuur and we also publish in these magazines ourselves. Scientific journals are hardly ever read by professionals.‘
The fact that the research is funded by the practitioners is of course a bonus, but I find it much more relevant that through working together with practitioners you can conduct research into pressing and relevant social issues, such as digitalisation, pandemics, climate change or migration.
What do you gain from this way of working? But also what does the university and society get out of it?
‘I think this way of working leads to an effective knowledge transfer to society. As a researcher, you form a completely different picture of your research field, which I would not have had if I had continued to conduct research from my ‘ivory tower perspective’. I very deliberately call what we do ‘a method’, because it systematic approach to collecting and analysing evidence. And it makes it much easier to connect science,practice, and education. The fact that the research is funded by the practitioners is of course a bonus, but I find it much more relevant that through working together with practitioners you can conduct research into pressing and relevant social issues, such as digitalisation, pandemics, climate change or migration. You need that transdisciplinary way of working to be able to use the knowledge from different sectors of society and to bring about change. But our teaching is changing too, because we adapt what we explain to students. With our knowledge of the professional field, we can prepare students much better for what they will encounter in the real world.'
UU continuing education is based on subjects that we as a university know something about and think are relevant. The starting point is different with the way you work. Do you have any advice on this point?
‘The in-company education of the Utrecht Data School that we offer through our own network works very well; better than offering the course via the UU website. If we offer an existing course via our website, we have to invest a lot of time and energy to attract enough participants. But the in-company courses are often linked to tools that we develop in and with the field. Think of our Fundamental Rights and Algorithms Impact Assessment (FRAIA). Government organisations want their employees to learn how to apply it. And we can offer this in special train-the-trainer courses. It also contributes effectively to the dissemination of FRAIA.'
‘To improve the education for professionals a university could get into discussion with someone from human resources, or maybe hire such a person. Someone who knows what the training budgets of large companies are, and how you can motivate these companies to offer your services to their employees. You have to consider for yourself how far you want to go in adapting education to the needs of external partners. Those are strategic choices you have to make.’