Events

Max Planck Institute for Innovation and Competition, Munich, Room 313

In this paper we investigate the effect of an unexpected increase in the availability of scientific articles on the follow-on scientific usage of the knowledge incorporated. We focus on the launch of Sci-Hub, a Kazakhstan-based website that provides free access to scientific literature, gathering data from the first three months of activity of the website (from September to December 2011). Then we link downloaded scientific articles to their corresponding bibliographical information retrieved from Web of Science. Finally we reconstruct the entire flow of citations pertaining to these scientific articles to measure the effects of a reduction in their access costs on their rate of usage within the scientific community. Our main hypothesis is that reducing the cost of accessing scientific knowledge lead to higher rates of knowledge usage by the scientific community. The introduction of Sci-Hub induced a large increase in citations to downloaded articles coming from scholars located in developing countries. This effect is persistent across article cohorts. As expected, the effect is absent when we consider citations whose scholars are located either in European developed countries or in the United States.

Contact Person: Michael E. Rose

Max Planck Institute for Innovation and Competition, Room 313

Max Planck Institute for Innovation and Competition

Please find further information here:

https://www.southcentre.int/call-for-papers-september-2019/

Max Planck Institute for Innovation and Competition, Munich, Room 313

It is by now well-documented that performance pay has positive effort and selection effects in routine, easy to measure tasks, but its effect in knowledge creation is much less understood. This paper studies the effect of performance pay on knowledge creation through effort and selection effects using the introduction of performance pay in German academia as a natural experiment. To this end, I consolidated information from various, unstructured data sources to construct a data set that encompasses the affiliation history and publication records of the universe of academics in Germany. The performance pay reform introduced attraction and retention bonuses, as well as relatively weaker on-the-job performance bonuses that take effect at a later point in time. I estimate the pure effort effect of these performance pay incentives in a difference-in-differences framework, comparing changes in research productivity of a treated cohort of academics, who receive performance pay because they started their first tenured position after the reform, with a control cohort that receives flat wages because they started their first tenured position just before the reform. I find a positive effort effect of performance pay that is economically large; amounting to a 12 to 16% average increase in research productivity. This increase manifests itself most robustly as an increase in research quantity and persists for a number of years. The effort response is strongest and most robust for less productive academics, with increases in pure quantity as well as quality-adjusted research output, while the average impact of the work of top quartile academics decreases. Performing textual analysis on paper abstracts to construct novelty and impact metrics, I find that the novelty of the work of top quartile academics declines. This work however does find more follow-on research in subsequent papers in the same field and is thus more impactful. I estimate the selection effect by analyzing the rate at which academics of different productivity levels switch to the performance pay scheme. I use the fact that the old and new wage schemes compare differently for academics at different ages, which gives rise to selection incentives that are inversely related to age. Exploiting this variation in a difference-in-differences framework, I find that more productive academics are more likely to select into performance pay. Hence, performance pay increases research output in academia through both effort and selection effects. However, because the effort effect is strongest for relatively less productive academics, while relatively more productive academics select into performance pay, the selection effect partially counteracts the impact of the effort effect.

Contact Person: Marina Chugunova

On 28 and 29 November the “Artificial Intelligence & Intellectual Property Conference” will take place in Singapore. The event, organized by the School of Law of the Singapore Management University (SMU) together with the Max Planck Institute for Innovation and Competition and the Law Faculty of the Chinese University of Hong Kong (CUHK), brings together academics, policy makers, lawyers and practitioners. The program focuses on the immediate legal impact of AI in different countries. The conference will discuss not only theoretical approaches, but also the implementation of new technology-oriented laws and guidelines for regulators considering an amendment to their IP laws.

See Conference Website

Max Planck Institut for Innovation and Competition, Munich, Room 313

The increasing use of algorithms in legal and economic decision-making has led to calls for a “right to explanation” to be given to the subjects of automated decision-making. A growing literature in computer science has proposed a vast number of methods to generate such explanations. At the same time, legal and social science scholars have discussed what characteristics explanations should have to make them legally and ethically acceptable. These debates suffer from two shortcomings. First, very little connection exists between these two strands of literature. Second, we do not know what effects such explanations would have on the behavior of decision subjects and on their perception of decision-making algorithms. In this field experiment, we aim to address these gaps by empirically testing how different types of explanations affect the subjects’ attitude towards decision-making algorithms. Distilling various factors that constitute a good explanation of algorithmic decision-making, we collect data on which factors are useful to decision subjects: local or global explanations, explanations which are selective, contrastive and/or are displayed as conditional control statements versus correlations. In the setting of a scholarship awarded by a machine learning algorithm to promising students, our experiment thus investigates which kind of explanations can lead to increased acceptance of algorithmic decision-making.

Contact Person: Dr. Marina Chugunova

The phenomena connected to the development of smart urban mobility solutions raise issues in various legal fields pinpointing the interaction between technology, law, and public policy. This complex relationship can be tackled from different angles of analysis: the public sector perspective, the private sector perspective, and the citizens’ perspective.

Max Planck Institute for Innovation and Competition, Munich, Room E10

Professor Wang introduces a theory of intelligence, a formal model of the theory, and a computer implementation of the model. He takes “Intelligence” as the ability of adaptation under insufficient knowledge and resources. Non-Axiomatic Reasoning System (NARS) is a formal model realizing this theory, which has been implemented in an open source project OpenNARS. The system shows many properties observed in the human mind. Practical applications of this technique are also under development.

Max Planck Institute for Innovation and Competition, Munich, Room 313

Strategic behavior can not only affect effort in contests but also undermine their selection function. We investigate two forms of strategic behavior of contestants with heterogeneous ability in large contests: Sabotage and self-promotion. We test predictions from a simple theoretical model in a large dataset of more than 38 million peer-ratings by 75,000 individuals. We find a) that strategic behavior influences outcomes in 25% of close contests, b) that self-promotion is the dominant form of strategic behavior of low-ability contestants, and c) that high-ability contestants are both culprits and targets of sabotage. We leverage two natural experiments to rule out alternative explanations.

Contact Person: Klaus Keller, M.A.