“If you want to publish at the highest level, there’s no getting around open science.”

Lukas Fink on his experiences with Open Science

What does Open Science mean to you personally, and why is it particularly important in the early stages of your career?

LF: For me, Open Science means making the entire research process transparent and communicating results openly. This strengthens trust and facilitates access to knowledge. I already noticed the practical implications of this during my master’s degree. In many empirical seminars, we first had to reproduce the results of a reference paper. Without access to data or code, this was almost impossible. It was completely different in my master’s thesis, in which I dealt with a new econometric method: there, I deliberately based my application examples on studies whose data and code were publicly available. This enabled me to better understand the existing analyses, comprehend the results more easily and continue my work much more efficiently. This experience showed me how important openness is for knowledge transfer. That’s why I want to consistently apply open science in my own research.

When you were asked to reproduce existing studies during your studies, was that a didactic approach by the lecturers to give you that ‘aha’ moment, or did you choose the studies yourself?

LF: Reproductions and replications are common in many seminars, even if they are rarely referred to as such. Lecturers usually suggest studies based on well-known data sets as references because they make it easier to develop new topics – for example, with a different target variable, a different time period or a different method. However, this was not systematically taught as a “replication course”; it is more of a pragmatic approach. In my master’s thesis, on the other hand, I made the selection myself and deliberately chose only studies that had openly accessible data and code. This allowed me to understand the existing analyses in detail, reproduce the results and then test their robustness against the new econometric method. For me, this was the first real step towards systematic reproductions.

Was there a particular moment that sparked your interest in open science?

LF: My studies were definitely the first point of contact. That’s where I realised how difficult reproductions are without data and code. Another important moment was my collaboration with my supervisor, Professor Jan Marcus. In a joint project, we investigated how often studies based on the Socio-Economic Panel provide replication code. During this project, I also took a more fundamental look at open science practices and understood that transparency and traceability not only help me personally, but also benefit research as a whole. This had a lasting impact on my interest in the topic.

How do your colleagues react to your openness in the research process? Is Open Science now taken for granted?

LF: The response has been overwhelmingly positive. There was a lot of interest in the paper I co-authored with Jan Marcus – and some surprise that only about six per cent of studies provide replication material. Many people realise that openness brings great added value, especially when it comes to the availability of data and code. Attitudes towards pre-registration are more nuanced. Some see it as a restriction to have to commit to analysis methods and data steps at an early stage. Pre-registration is now largely standard practice, especially in experimental research, whereas it is still rarely used in observational studies. There is also concern that deviations could be interpreted negatively in peer review.

What role do professional associations such as the Verein für Socialpolitik play in promoting Open Science? Should they primarily address the younger generation or rather older, sceptical researchers?

LF: Professional associations are important for raising awareness of the issue and reaching a broad audience. Conferences, workshops and panels create spaces for exchange, and through their journals, professional associations can set standards by not only promoting open science practices, but also making them mandatory.

Could one say that professional societies primarily raise awareness, while journals ultimately ensure implementation by setting specific requirements?

LF: Yes, that division of tasks describes it quite well. Professional societies can raise awareness and provide guidance, while journals then set binding standards. But raising awareness alone can also have an impact. One example is the American Society of Health Economics: together with its journals, it has called for zero results to be taken into account and not devalued. Studies, such as those by Brodeur and co-authors and a paper by Thiemo Fetzer, show that this call has actually led to more null results being published in Health Economics. Such signals can therefore have measurable effects – even without immediate formal commitments.

When it comes to raising awareness, there are two distinct approaches: negative arguments – such as replication crises and alarming findings – or positive arguments, i.e. the concrete advantages of Open Science for the discipline and researchers. Which do you think is more effective?

LF: One example is the 2015 study by the Open Science Collaboration in psychology, which showed that many experiments could not be successfully replicated. These findings were widely reported, including by The Economist. The term “replication crisis” arose in this context and led to many psychology journals introducing more binding data and code policies from the mid-2010s onwards. Our own research also shows that the availability of replication material has increased massively since then. At the same time, I think it is crucial to emphasise the positive aspects. In discussions, especially with more experienced colleagues, it is often more convincing to point out the benefits of open science practices for one’s own work, such as working in a more structured manner and being able to easily understand and reproduce one’s own analyses later on, for example in the revision process. In addition, open material also helps other researchers to better understand one’s own work. Personally, I therefore prefer to emphasise the advantages rather than portraying the state of science as a crisis.

What is actually preventing the discipline from being much further ahead?

LF: I wonder about that too, but I do see movement. More and more journals, not just the top ones, have now introduced mandatory data and code policies. In addition, there are new roles such as data editors, who not only check the reproducibility of studies prior to publication, but also provide guidelines, blogs and training courses. There are also initiatives such as the Institute for Replication, which provide information and disseminate standards. As a result, the topic has gained significantly in visibility and momentum in recent years.

What would an ideal “open economics world” look like for you – your desired working environment for the next few years?

LF: In an ideal environment, data and code would be shared as a matter of course, and researchers would also receive recognition for this. Practices such as pre-registrations or pre-analysis plans would be established and seen as a mark of quality. My experience so far shows that this leads to a more structured approach to projects, clear agreements within the team at an early stage and more efficient work overall. This is particularly valuable for early-career researchers, as they rarely specify the basic research idea, especially in their first projects, and are thus involved in intensive discussions at an early stage. It would be crucial for this benefit to be more strongly internalised in the community and for open science practices to be rewarded in a way that is relevant to careers.

What experiences would you describe as personal game changers – things you have done that you would wholeheartedly recommend to others?

LF: One formative experience was when researchers who had read our study on the availability of replication code contacted us and wanted to use part of it themselves. We were able to refer them directly to our complete replication package, which we had uploaded to Zenodo. Our colleagues thanked us and announced that they would cite us. That was very motivating. My work on open science also led to a collaboration with colleagues from RWI. There, we want to systematically carry out replications on particularly policy-relevant research questions. This shows that openness not only creates transparency, but also opens up new contacts and opportunities for cooperation. Furthermore, I have found it beneficial to systematically document data and work steps from the outset. This not only makes my own work easier, but also facilitates reuse by others.

Does feedback from the professional community play a special role for you, especially as a young researcher?

LF: Yes, definitely. It’s a good feeling when your own work is noticed and taken seriously. One example was the request for our data, which showed that someone was really engaging with our research. The positive response to our paper, for example through its dissemination via Bluesky, was also motivating. The same goes for feedback at conferences: especially when you are working in a field of research that is not yet well established, it is encouraging to hear that your contributions are considered relevant and important.

Let’s imagine that at the annual conference of the VfS, you meet someone who says, “I don’t have a supervisor like Jan Marcus, but I want to get into open science.” What would you advise – peer to peer?

LF: My advice would be to talk to your supervisor early on and be open about your plans for open science practices. In most cases, there is no serious resistance, especially when it comes to publishing code and data. Often, it is the junior researchers who write the code and can therefore ensure its traceability and reproducibility. It is important to emphasise the advantages: if, for example, you work towards a replication package that is intended for sharing right from the start, this usually also facilitates teamwork.

And what would you say to sceptics? What are your three strongest arguments for Open Science?

LF: My first and strongest argument would be that top journals now often require replication packages or even pre-registration of experiments. Anyone who wants to publish in these journals can hardly avoid open science practices. Secondly, reproducible code improves your own workflow. You work in a more structured way, can understand results more easily and are also more efficient in collaborations. These skills are of great value not only in science, but also in the private sector. Thirdly, Open Science is a public good. By making data, code and methods transparent, we strengthen trust in research – within the scientific community and in society as a whole. If we want to influence society, we need its trust. We must not lose sight of this goal.

Very nice, that sums it up: future skills. Your first two arguments show how important it is to invest in your own skills and in good collaboration potential in order to be a strong partner in research. A good conclusion. Thank you very much for the interview.

*The interview was conducted on 8 September 2025 by Dr Doreen Siegfried.
This text was translated on 29 October 2025 using DeeplPro.

About Lukas Fink:

Lukas Fink has been a research assistant at the Chair of Applied Statistics at the Free University of Berlin since April 2023. He completed his master’s degree in public economics at the Free University of Berlin in 2023. Prior to that, he obtained his bachelor’s degree in economics from the University of Duisburg-Essen in 2020. His research interests lie particularly in the areas of economics of science, public economics and issues of reproducibility and replicability.

Contact: https://www.wiwiss.fu-berlin.de/fachbereich/vwl/angewandte-statistik/Team/mitarbeiter_innen/fink/index.html

Bluesky: @lukasfink.bsky.social

GitHub: https://github.com/Finkovicius

ORCiD: https://orcid.org/0009-0002-8786-0988

to Open Science Magazine