Research News

Citizen Science thrives on the interaction between citizens and researchers. The interest in cooperation is steadily growing. Well-known projects in the environmental area include bird counting and bee observation.

Since December 2021, the MPI for Innovation and Competition has now been cooperating with the German Association for Computer Genealogy (CompGen) in a data project to record the annual directories of publications at German universities and higher education institutions. The directories, which were published between 1885 and 1987, first by the Royal Library in Berlin and later by the German Library in Leipzig, cover 103 volumes. They list mainly dissertations and postdoctoral theses written at German universities and higher education institutions. Afterwards, the directories were discontinued in this form. A digital continuation failed.

For citizens who conduct genealogical or family research, the lists, some of which contain rich biographical information, are interesting because they hope to meet ancestors, bearers of the same name or people from their town or region. Birgit Casper, who is working on the project, reports on her motivation for collecting the data: “I know two doctors in my family. Of one, born in 1891, I know pretty much where he studied and that he submitted his dissertation ‘On cases of poisoning with American worm seed oil’ to the medical faculty in Rostock in 1920. Of the other, born in 1892, I only know where he practiced medicine as of 1924. I do not know where he studied, nor when and on what subject he did his doctorate. Here I am waiting for the corresponding volume.”

For scientists, the lists are intriguing because they provide a complete overview of researchers who were educated at German universities since 1885 and some of whom were internationally important. Since German universities were internationally leading in almost all disciplines at the turn of the century, the project promises particularly interesting insights. We find the dissertations of numerous later Nobel Prize winners, such as Walther Nernst, who received the Nobel Prize in Chemistry in 1920 and was on the board of directors of the Kaiser Wilhelm Institute (KWI) for Physics, as well as Werner Heisenberg, who later gave his name to the subsequent MPI for Physics, and also Maria Goeppert-Mayer, who was awarded the Nobel Prize in Physics in 1963 as the first female German Nobel Prize winner.

The first regular doctorate for a woman will also be found in the lists. In fact, women were severely underrepresented at first. Only a few were allowed to earn a doctorate before 1900, and only with special permission. It was only between 1901 and 1908 that the German states successively admitted women to their universities. The right to pursue doctoral studies, however, was awarded by the faculties themselves. A systematic recording of all dissertations will thus generate a complete overview of when, at the latest, women were allowed to earn a doctorate at which universities and faculties. The right to habilitation – the path to professorship – was given to them even later: Here, too, the lists can help shed light on the situation.

How does the collaboration between science and amateurs work in the project?

To find volunteers who want to work on the project, CompGen publishes calls and updates on Twitter and in the blog on the Compgen website. On a special wiki page for the project, volunteers can register, learn about the editorial guidelines, and start editing data right away.

Michael E. Rose, Senior Research Fellow at the MPI for Innovation and Competition, who leads the project and is active in the field of Science of Science, is gradually scanning the directories.

Then the lists are captured with a text recognition program and roughly segmented: What are first name, last name, title of the dissertation, the date of the defense, other details? The volunteers use the infrastructure provided by the association (input mask and data repository) to proofread the entries and add to them manually. The entered records are immediately available for search queries. So far, seven annual directories have been processed. After completion of the project, the lists, which are interlinked, for example, to the German National Library, Wikipedia, and Scopus, a multidisciplinary abstract and citation database for research literature, will be publicly available as research data.

One of the best-known personalities recorded so far is Fritz Haber, who, as founding director, headed the KWI for Physical Chemistry and Electrochemistry in Berlin for 22 years, which is now named after him. His dissertation, „Ueber einige Derivate des Piperonals“ (On some derivatives of piperonal), a fungicidal fragrance, is found in volume VI (1890/91). Fritz Haber received the Nobel Prize in 1919, awarded for the year 1918, for his research on the catalytic synthesis of ammonia, i.e., in a different field of research from his dissertation.

Max von Laue, on the other hand, who completed his doctorate with Max Planck in 1903 „Über die Interferenzerscheinungen an planparallelen Platten“ (On the interference phenomena in plane-parallel plates), continued to pursue the research begun with his dissertation – until he was awarded the Nobel Prize in 1914 for his work on X-ray interferences.

However, not all doctoral graduates were able to receive the recognition they deserved. Hilde Mangold’s research in the field of embryology led to a Nobel Prize in 1935 for her doctoral advisor Hans Spemann, who was director at the KWI for Biology in Berlin-Dahlem during the First World War. Mangold herself died in a fire shortly after defending her dissertation in 1924. After all, the prize-winning discovery, the Spemann organizer, is sometimes called the Spemann-Mangold organizer.

Due to their depth of detail and completeness, the data digitized in the project allow for numerous exciting research questions. Can we read problems of an era from law dissertations? How do the demographics and social origins of doctoral students change over time and at individual universities? Who were the women who were the pioneers in earning a doctorate? What is the relationship between dissertations and patent activity?

However, before that, the dataset must be completed, and every hand and pair of eyes is still needed to accomplish this. More information under https://wiki.genealogy.net/Hochschulschriften.

This second Position Statement, which follows the Institute's Position Statement of 7 May 2021, reflects on the legal and practical implications of the Ministerial Decision in view of the ultimate goal of overcoming the COVID-19 pandemic. A particular focus here is on TRIPS flexibilities relating to compulsory licensing of patents.

Position Statement of 5 July 2022 on the Decision of the WTO Ministerial Conference on the TRIPS Agreement adopted on 17 June 2022

Directly to the Position Statement (PDF).

As technology-assisted decision-making becomes more prevalent, it is important to understand how the algorithmic nature of the decision-maker influences how affected people perceive these decisions. The application of algorithmic aids spans from prediction decisions of various kinds, for example, whom to hire, and what salary to offer, to moral decisions with no objectively correct solution, such as how to distribute a bonus within a team fairly.

The authors Marina Chugunova and Wolfgang J. Luhan (University of Portsmouth) use a laboratory experiment to study the preference for human or algorithmic decision-makers in redistributive decisions. Redistributive decisions can be seen as a type of moral decisions, where the definition of correct or fair depends on the observer’s personal ideals and beliefs. In particular, the authors consider whether an algorithmic decision-maker will be preferred because of its unbiasedness. Defining which decision-maker is preferred and whose decisions are perceived to be fairer can potentially improve the acceptance of decisions or policies, and with it, the compliance.

The Experiment

In the experiment, the main aim was to create a situation where participants’ preference for either a human or an algorithmic decision-maker to redistribute income was observable. First, participants individually earned their initial income by completing three tasks. The three tasks mimicked three potential determinants of income that are central to major fairness theories: luck, effort and talent. Then, the players were matched into pairs and had to choose a decision-maker: either an algorithm or a third party human. The decision-maker decided how to redistribute the total earnings of the pair between the two members. To test the role of bias, a laboratory-induced source of potential discrimination for the human decision-maker was introduced. Then, the participants learned the decision and had to report their satisfaction and their rating of how fair a particular redistribution decision was.

The Findings

Contrary to previous findings, the authors find that the majority of participants ‒ with over 60% ‒ prefer the algorithm as a decision-maker over a human. Yet, this is not driven by concerns over biased decisions of a human. Despite the preference for algorithmic decision-makers, the decisions made by humans are regarded more favorably. Subjective ratings of the decisions are mainly driven by own material interests and fairness ideals. As far as fairness ideals are concerned, the players in the experiment show a remarkable flexibility: they tolerate any explainable deviation between the actual decision and their own ideals. They are satisfied and consider any redistribution decision that follows fairness principles to be fair, even if it does not correspond to their own principles. Yet, they react very strongly and negatively to redistribution decisions that do not fit any fairness ideals.

The Conclusion

The results of the study suggest that even in the realm of moral decisions algorithmic decision-makers might be preferred over human decision-makers, but the actual performance of the algorithm plays an important role in how the decisions are rated. To “live up to the expectations” and increase the acceptance of these AI decisions, the algorithm has to consistently and coherently apply fairness principles.

Directly to the publication of the study:

Marina Chugunova, Wolfgang J. Luhan
Ruled by Robots: Preference for Algorithmic Decision Makers and Perceptions of Their Choices
Max Planck Institute for Innovation & Competition Research Paper No. 22-04

The idea of limiting the power of digital giants is not new. Rules punishing abuse of dominant position or unfair business practices already exist. What's about to change?

The Digital Markets Act, or DMA for short, which is currently being negotiated in non-public consultations between the EU Parliament, Council and Commission, aims to provide a uniform EU-wide legal framework for digital markets in order to proactively prevent distortions of competition and stop certain business practices.The aim is also to facilitate the application of the law and speed up the process.

Why are the existing rules insufficient?

So far, especially in antitrust law, we have only ex-post rules: these only take effect after an infringement has occurred or is suspected. As a result, it takes a very long time before the behavior is sanctioned. In some cases, the contested acts took place up to ten years ago and the legal proceedings have still not been concluded. The result is that antitrust law is too slow because it takes action reactively.

What should change?

The basic concept of the DMA is upfront regulation: it will impose a mandatory code of conduct on major online service providers. With this in mind, it contains over a dozen directives and prohibitions, formulated in more or less specific language, with which these service providers are required to comply a priori. Failure to comply could result in severe sanctions. For example, the EU Commission has the power to fine service providers up to 20 per cent of their annual global turnover.

We'll come to the new rules in a moment. But first: Who does the planned regulation actually apply to?

According to the European Council's first press release, a service provider will be considered a "gatekeeper" if the parent company generates an annual turnover in excess of €7.5 billion or has a market value of more than €75 billion in the EU. The service also has to have had over 45 million end users per month and 10,000 commercial users within the EU per year over the past three years.

The respective service provider also has to offer one or more core services, such as social networks, search engines, browsers, messenger services, or virtual assistants, in at least three EU member states. However, networked TVs are not included.

Which obligations will gatekeepers have in the future?

So far, there are 18 obligations, but the list of obligations may be larger after the negotiations. Particularly relevant, for example, are prohibitions on tying registration for one service to registration for another, or prohibiting commercial users from offering their products and services on third-party platforms at different prices and conditions. In addition, operating systems such as Google's Android or Apple's iOS would have to allow app stores other than their own to be installed on smartphones. Self-preferential practices are to be explicitly prohibited, for example by offering consumers preferential treatment for the company's own services, thereby closing the market to smaller or new competitors. In the future, it will also no longer be permitted to combine and use data for certain purposes, as Facebook is accused of doing.

 

The new interoperability requirement is also of particular relevance to consumers. For example, messaging services that have been classified as gatekeepers will be required to permit their users to exchange messages and make video calls between different platforms (e.g., Whatsapp and Threema).

Sounds like Lex Google, Lex Apple, Lex Facebook ...

Indeed, many of the specific provisions are based on cases in which the EU Commission and the national competition authorities have taken legal action against these companies. Individual laws cannot exist, but it is of course possible to define abstract criteria for market activity in the EU in order to prevent as much problematic behavior as possible.

Certainly, it will hit the U.S. tech giants in particular with their services at the beginning: Apple's and Google's Appstore, Microsoft's Bing in addition to Google's search engine, Facebook and Whatsapp or Amazon Marketplace. The cloud computing segment of Amazon, for example, which is important for B2B and accounts for the largest share of the corporate profit, will also be affected.  

However, the extent to which the regulations will apply to European platform services such as Booking or Zalando remains unclear.

Can the regulation make the digital markets more open and fairer?

The crucial point is how the regulations will be applied and enforced in practice. So whether the agreed procedure will work as intended – including, in particular, the dialogue foreseen between the EU Commission and the “gatekeepers” concerning the regulatory measures – will be clarified in due course. Another open question is whether the envisaged collaboration between the national authorities and the EU Commission will be effective. It also remains to be seen which role the national courts will ultimately play in enforcing the DMA regulations.

The EU Commission is expected to play a significant role. What is it?

Although it has been controversial among the member states, the EU Commission is to be given a very central role. The DMA permits the Commission to order a wide variety of measures against gatekeepers. These can be behavioral measures, i.e. that a company may no longer do something specific or must do something, and structural measures such as the separation of entire business units. At the same time, the Commission is supposed to monitor compliance with the orders. All this requires the EU Commission to create numerous additional positions with specific expertise. Indeed, the EU Commission will be taking on a completely new role as a regulatory authority, and it will be interesting to see if it will ultimately be able to stand up to the big tech companies on an equal footing.

Assuming everything goes according to plan and the European Parliament and Council give their official consent, the new regulation will come into force in 2023. Can all these innovative provisions be implemented immediately?

That remains to be seen. The problem is that many technical issues that are legally relevant are not solved, but only postponed. The fine-tuning is done by means of so-called Delegated Acts, which have yet to be issued by the Commission. In reality, this can take years. One example: Until now, Deutsche Bahn has not been obliged to make real-time information about its trains available to other transport information services such as Google Maps. The relevant EU Directive governing this has been in place since 2010, but the legal details came in 2017 in a Delegated Act from the Commission. The rules were not incorporated into the German Passenger Transportation Act until 2021, and it is still not clear to what extent Deutsche Bahn will have to provide real-time information.

What else is needed?

The effectiveness of enforcement by the public sector, in this case by the EU Commission, always depends on the political impetus. And priorities can change quickly in this regard. The current Commission, with Margrethe Verstager as Digital Commissioner, has the DMA high on its list, so the rules have come surprisingly fast. But it doesn't guarantee that if there is a new Commission with a new agenda, it will continue with the same verve.

The interview was conducted by Michaela Hutterer

Personal website of Heiko Richter

A long-running discussion about copyright in academic publishing has shown the role of copyright and its dysfunctional effects.

The interests of commercial publishers and other information providers differ from those of academic authors, with the former usually pursuing a strategy of profit maximization, while the latter want to ensure broad access, open and timely dissemination, and reuse of scientific results. Moreover, third parties usually fund research, so academic authors do not primarily rely on income from publications – researchers publish primarily to enhance their reputation and advance their careers.

In this context, the contribution by Valentina Moscon (Max Planck Institute for Innovation and Competition) and Marco Bellia (Università Cattolica del Sacro Cuore) draws attention to new models that promise a fairer and more efficient scholarly publishing system. After reviewing the legal background in Italy, Germany, and the United States, the authors consider various possible interventions, some of which have already been adopted at the national level. These measures may be private interventions, such as university contracts and policies, or public, i.e., legislative interventions. The latter include measures outside or inside the copyright system.

“International Instrument” as a model for a fair solution

The authors conclude that the best solution is to redefine the boundaries of copyright by broadening the scope of permitted uses while defining them more precisely. This would lead to a more balanced functioning of the academic publishing system. One proposal in this direction comes from a group of copyright experts, including Valentina Moscon, who have drafted the International Instrument on Permitted Uses in Copyright Law. This instrument, conceived in the form of an international treaty, aims to create a more balanced system for the scope of international copyright protection. Among other provisions, it contains explicit rules for permissible uses in academia, including uses in the context of research, data analysis, educational purposes, and for the processing of works by libraries, museums, and archives.

To the publication:

Marco Bellia, Valentina Moscon
Academic Authors, Copyright and Dissemination of Knowledge: A Comparative Overview
Max Planck Institute for Innovation & Competition Research Paper No. 21-27

The authors establish the importance of physical capital in knowledge production. To this end, they exploit adverse events (explosions, fires, floods, etc.) at research institutions as exogenous physical capital shocks. Scientists experience a substantial and persistent reduction in research output if they lose specialized physical capital, that is, equipment and material they created over time for a particular research purpose. In contrast, they quickly recover if they only lose generic physical capital. Affected scientists in older laboratories who presumably lose more obsolete physical capital, are more likely to change their direction of research and recover in scientific productivity. These findings suggest that a scientist’s investments into their own physical capital yield lasting returns but also create path dependence in relation to research direction.

The study suggests that science and innovation policy should give more consideration to the role of physical capital in knowledge production.

Directly to the publication by
Stefano Baruffaldi and Fabian Gaessler
The Returns to Physical Capital in Knowledge Production: Evidence from Lab Disasters
Max Planck Institute for Innovation & Competition Research Paper No. 21-19

In particular, the position statement critically views the arguments to do so provided in a recent decision by the Australian Federal Court and highlights that they are based on questionable assumptions, first, regarding the technical capabilities of AI and, second, concerning the ostensibly positive impact of such recognition on innovation. Overall, it is emphasised that the question that should be addressed in the first place is whether inventions allegedly generated by AI should merit patent protection at all.

Position Statement