Promotion of reproducibility and replication in economic research
ZBW symposium discusses research transparency in economics
Photo: David Außerhofer
Research transparency is a central element of scientific integrity and essential for creating trust in research results. Trust in science is essential if scientific findings are to be accepted by society and used as a basis for decision-making. However, the Science Barometer 2023 from “Science in Dialogue” shows that although trust in scientific institutions remains high at 56%, worrying trends are becoming apparent in which individual social groups are becoming increasingly sceptical of research results and scientific processes. The proportion of respondents who state that they (tend to) have no trust in science and research has risen slightly compared to the previous year and stands at 13 per cent in the current 2023 survey.
Research transparency refers to the disclosure and traceability of the methods, data and processes used to produce scientific results. Transparency is of particular importance as it enables the reproducibility and verifiability of research results. In times of increasing digitalisation and data availability, the demand for openness and transparency in science is becoming even more urgent.
Research transparency plays a special role in economic research, as the results of economic studies often have far-reaching social and political implications. Whether it is macroeconomic forecasts, studies on markets or economic policy recommendations – transparent research is crucial in order to assess the validity and relevance of the findings. This is all the more true as many economic models are based on complex data whose replicability is not always guaranteed.
ZBW Symposium: Promoting reproducibility and replication in economic research
On 27 April 2023, the ZBW – Leibniz Information Centre for Economics organised a symposium that explicitly addressed the topics of reproducibility and replication in economic research. Under the direction of Prof Dr Marianne Saam, Professor of Digital Economics at the ZBW, the ZBW’s first Open Science Symposium took place, initiating a broad discussion on the importance of openness in economic research.
The reproducibility of scientific studies – i.e. that other researchers can achieve the same results based on the original data and methods – is one of the basic requirements for reliable research. This poses a particular challenge in economic research, as many analyses are based on proprietary data that is often difficult to access. At the same time, the demand for replication – i.e. the independent repetition of studies – is becoming ever louder in order to recognise possible errors and strengthen confidence in research results.
Four national and international speakers shared their experiences in the field of research transparency:
Melanie Schienle, Professor of Statistical Methods and Econometrics at the Karlsruhe Institute of Technology, showed in her presentation that a collaborative and combinatorial approach for short-term predictions in real time is promising. The Covid19-Nowcasthub was developed at her institute, a platform on which an ensemble for hospitalisation incidence is calculated from several models.
Dr Lars Vilhuber, Data Editor of the American Economic Association, presented the AEA’s policy for publication in its journals, which has been in force since 2019. It stipulates that articles will only be published if the analysis and access to the data and code used is clearly documented and not exclusively reserved for the authors.
In his presentation, Joachim Gassen, Professor of Accounting and Auditing, came to the conclusion that the reproducibility and replicability of research are key components of scientific progress, but that Open Science has its real impact where others can build on the results and methods of previous research.
Professor Jörg Ankel-Peters from the RWI – Leibniz Institute for Economic Research and founding member of the Institute for Replication asked in his presentation whether transparency standards are enough. He criticised the lack of a replication culture and the fact that only 1 to 3 percent of published articles are policing replications, i.e. they directly address and challenge a previously published paper. This is because a replication can only develop its regulating effect through such a direct approach.
The importance of open science for economic research
At the symposium, there was intensive discussion about how open science practices can contribute to promoting research transparency. The symposium participants discussed that publishing datasets, code and detailed methods not only improves the traceability of results, but also facilitates collaboration and exchange between researchers. However, the researchers also identified challenges. Many economic researchers face structural and institutional barriers when it comes to disclosing their data. These include copyright and data protection restrictions, as well as competitive pressure within the scientific community. Nevertheless, the participants emphasised the importance of a cultural shift towards greater transparency, in particular through the stronger anchoring of open science practices in academic education and through institutional incentives.
The text was written on 22 October 2024.
This text was translated on 20 January 2025 using DeeplPro.