Universitätspublikationen
Refine
Year of publication
Document Type
- Doctoral Thesis (35) (remove)
Has Fulltext
- yes (35)
Is part of the Bibliography
- no (35)
Keywords
- Geldpolitik (3)
- Deutschland (2)
- Dogma (2)
- Economic Ethics (2)
- Eucken, Walter (2)
- Freiburger Schule <Wirtschaftswissenschaften> (2)
- Geschichte (2)
- International Macroeconomics (2)
- Makroökonomie (2)
- Monetary Policy (2)
Institute
- Wirtschaftswissenschaften (35) (remove)
This cumulative dissertation contains four self-contained chapters on stochastic games and learning in intertemporal choice.
Chapter 1 presents an experiment on value learning in a setting where actions have both immediate and delayed consequences. Subjects make a series of choices between abstract options, with values that have to be learned by sampling. Each option is associated with two payoff components: One is revealed immediately after the choice, the other with one round delay. Objectively, both payoff components are equally important, but most subjects systematically underreact to the delayed consequences. The resulting behavior appears impatient or myopic. However, there is no inherent reason to discount: All rewards are paid simultaneously, after the experiment. Elicited beliefs on the value of options are in accordance with choice behavior. These results demonstrate that revealed impatience may arise from frictions in learning, and that discounting does not necessarily reflect deep time preferences. In a treatment variation, subjects first learn passively from the evidence generated by others, before then making a series of own choices. Here, the underweighting of delayed consequences is attenuated, in particular for the earliest own decisions. Active decision making thus seems to play an important role in the emergence of the observed bias.
Chapter 2 introduces and proves existence of Markov quantal response equilibrium (QRE), an application of QRE to finite discounted stochastic games. We then study a specific case, logit Markov QRE, which arises when players react to total discounted payoffs using the logit choice rule with precision parameter λ. We show that the set of logit Markov QRE always contains a smooth path that leads from the unique QRE at λ = 0 to a stationary equilibrium of the game as λ goes to infinity. Following this path allows to solve arbitrary finite discounted stochastic games numerically; an implementation of this algorithm is publicly available as part of the package sgamesolver. We further show that all logit Markov QRE are ε-equilibria, with a bound for ε that is independent of the payoff function of the game and decreases hyperbolically in λ. Finally, we establish a link to reinforcement learning, by characterizing logit Markov QRE as the stationary points of a game dynamic that arises when all players follow the well-established reinforcement learning algorithm expected SARSA.
Chapter 3 introduces the logarithmic stochastic tracing procedure, a homotopy method to compute stationary equilibria for finite and discounted stochastic games. We build on the linear stochastic tracing procedure (Herings and Peeters 2004), but introduce logarithmic penalty terms as a regularization device, which brings two major improvements. First, the scope of the method is extended: it now has a convergence guarantee for all games of this class, rather than just generic ones. Second, by ensuring a smooth and interior solution path, computational performance is increased significantly. A ready-to-use implementation is publicly available. As demonstrated here, its speed compares quite favorable to other available algorithms, and it allows to solve games of considerable size in reasonable times. Because the method involves the gradual transformation of a prior into equilibrium strategies, it is possible to search the prior space and uncover potentially multiple equilibria and their respective basins of attraction. This also connects the method to established theory of equilibrium selection.
Chapter 4 introduces sgamesolver, a python package that uses the homotopy method to compute stationary equilibria of finite discounted stochastic games. A short user guide is complemented with discussion of the homotopy method, the two implemented homotopy functions logit Markov QRE and logarithmic tracing, and the predictor-corrector procedure and its implementation in sgamesolver. Basic and advanced use cases are demonstrated using several example games. Finally, we discuss the topic of symmetries in stochastic games.
Auszubildende sollen durch die Berufsausbildung u.a. die Kompetenz erlangen, berufliche Probleme zu lösen. Abschlussprüfungen dienen der Kompetenzerfassung, schriftlich-kaufmännische Prüfungsaufgaben bilden allerdings noch unzureichend Problemsituationen ab, deren Lösung Problemlösekompetenz erfordert. An der Erstellung von Prüfungsaufgaben sind auch Lehrkräfte kaufmännisch-beruflicher Schulen beteiligt. In der Arbeit wird untersucht, wie sie in der ersten und zweiten Phase der Lehrer*innenbildung auf das Erstellen problemhaltiger Aufgaben zu summativ-diagnostischen Zwecken vorbereitet werden. Hierfür werden Dokumentenanalysen zu beiden Phasen der Lehrer*innenbildung durchgeführt. Die Ergebnisse werden mittels einer Fragebogenstudie mit Studiengangsleiter*innen sowie Interviews mit Fachleiter*innen der Studienseminare gesichert. Um die Wahrnehmung angehender Lehrkräfte zu erfahren, werden Interviews mit Masterstudierenden der Wirtschaftspädagogik sowie Lehrkräfte im Vorbereitungsdienst (LiV) an kaufmännisch-beruflichen Schulen durchgeführt.
Durch die Vorstudien gelingt es, Optimierungsbedarfe in der Ausbildung von Lehrkräften kaufmännisch-beruflicher Schulen festzuhalten. Davon ausgehend wird ein Trainingskonzept begründet ausgewählt. Die Evaluation dessen erfolgt mittels einer quasi-experimentellen Studie mit Masterstudierende und LiV. Zur qualitativen Evaluation werden Interviews mit Teilnehmenden der Interventionsgruppe durchgeführt. Die Ergebnisse zeigen, dass die Teilnehmenden das Training als Intervention überwiegend positiv wahrnehmen und dieser, zumindest mit Blick auf das Erstellen von problemhaltigen Aufgaben, zu einem Lernzuwachs führt. Durch die bedarfsorientierte Intervention und dessen Evaluationsergebnisse wird ein Konzept vorgeschlagen, welches eine Lösung zur Deckung bestehender Optimierungsbedarfe bietet. Die Ergebnisse der Arbeit haben das Potential, langfristig einen Beitrag zur Verbesserung der Lehrer*innenbildung zu leisten und somit u.a. Assessmentaufgaben valider zu gestalten.
Intelligenz ist einer der stärksten Einzelprädikatoren für berufliche Leistungen. Hochintelligente Personen, sogenannte Hochbegabte, müssten daher generell hohe berufliche Leistungen erbringen. Dies steht jedoch im Widerspruch zu Darstellungen von Schwierigkeiten Hochbegabter im Beruf. Sollten die negativen Stereotype zutreffen, wäre dies aus ressourcenorientierter Sicht problematisch, da Organisationen das der Hochbegabung zugrunde liegende Potenzial der Mitarbeiter*innen nicht hinreichend nutzen. Die vorliegende Arbeit leistet einen Beitrag dazu, diese Diskrepanz aufzuklären, indem sie untersucht wie Hochbegabte ihre berufliche Situation (erfolgreich) gestalten. Nach einer Begriffsklärung und Darstellung des Zusammenhangs von Intelligenz und Beruf wird im ersten Beitrag der internationale Forschungsstand mithilfe eines Literature Reviews systematisiert. Aufbauend darauf werden im zweiten Beitrag mithilfe einer Fragebogenerhebung die überwiegend internationalen Beiträge durch eine deutsche Stichprobe und die bisher wenig betrachtete Gruppe von Hochbegabten gemäß Kompetenzdefinition ergänzt. Schließlich fokussiert der dritte Beitrag den Widerspruch, dass Hochbegabte generell erfolgreich und zufrieden im Beruf sind, jedoch individuell von Schwierigkeiten berichten. Dazu werden teilstrukturierte Interviews durchgeführt, um ein tieferes Verständnis für den Einfluss der Hochbegabung auf den Beruf sowie etwaige bestehende Unterschiede zu nicht-hochbegabten Personen zu gewinnen. Zudem werden Strategien erfasst, die Hochbegabte nutzen, um Unterschiede und eventuell daraus resultierende Schwierigkeiten zu bewältigen. Abschließend werden Implikationen für die Berufspraxis und Forschung diskutiert.
For private investors it is imperative to a) understand and define their own, individual risk preferences, b) assess their financial and demographic circumstances to determine the individual risk-taking potential, and c) form and maintain a well-diversified risky portfolio. The three chapters of my thesis each match one of these three tasks. \\ \noindent The first chapter of my thesis presents novel experimental evidence to test the existence of a potential projection bias in loss aversion, a significant determinant of investor preferences, thus matching task a). The second chapter is devoted to the determination of private investors' risk-taking potential based on their financial and socio-demographic circumstances, matching task b): In a large portfolio experiment, we examine the ability and heterogeneity of lay and professional advisors in matching investor demographics, such as age and income, with risky asset portfolio shares. The third and final chapter addresses the question on how to reach and maintain an efficient risky portfolio, therefore matching task c): It analyzes a decision support system for private investors that allows its users to simulate any arbitrary set of securities, and by reporting aggregated expected return and risk, to optimize their current portfolio.
This dissertation consists of four self-contained chapters in the overlapping fields of industrial organization and organizational economics on the topics pricing, careers and supervision. Each chapter is the result of an independent research project. The dissertation analyzes empirical research topics by exploring novel observational data sets. It sheds light on open questions in the economic profession by extending fundamental models on pricing in the first two chapters and by challenging conventional explanations and methods on careers and supervision in the last two chapters.
- Chapter 1:
The first chapter is based on joint work with Steffen Eibelshäuser. It models price competition among brick-and-mortar retailers with business hours. Specifically, we propose a dynamic model of intraday price competition featuring spatial differentiation and firm size heterogeneity. The model makes detailed predictions concerning equilibrium-pricing patterns. When spatial differentiation is high and consumers cannot easily switch between retailers, equilibrium prices are stable at oligopoly levels. When differentiation is low, equilibrium prices fluctuate in cycles. The shapes of the cycles depend on the level of differentiation and on retailers’ reaction times. When reaction times decrease, the number of price cycles increases. In a second step, we apply the model to the German retail gasoline market. Gasoline retailers have been using digital price tags for decades and fast-paced price competition with more than ten price changes per day is no exception. Our model has successfully predicted the emergence of an additional intraday subcycle in April 2017. Moreover, we were able to confirm several detailed predictions concerning the shape of equilibrium price paths and individual firm behavior. Finally, we calibrate the model using a generalized method of moments. The model fits the data remarkably well, with coefficients of determination ranging from 60% to 80%. We use the fitted model to evaluate a number of policy counterfactuals. Restricting price increases results in higher prices and decreased welfare, leading us to conclude that regulation of dynamic markets is highly complex and can easily backfire.
- Chapter 2:
The second chapter analyzes the price-matching policies of two gasoline retailers. Customers of these retailers that are able to provide evidence of competitors posting lower prices have the ability to claim price matches. As shown in the first chapter, the Edgeworth Cycle model rationalizes price fluctuations in the German gasoline retail market. To determine policy interactions in cycling markets, this chapter extends the classical Edgeworth Cycle model by price-matching. The model predicts that price-matching retailers post higher prices and initiate price increases. The price-consulted firm anticipates this strategy, posts lower prices, and provokes the implementing firm to restore the price more frequently. Consulted stations also anticipate earlier price restoration reactions from implementing stations and, thus, provoke restorations earlier. This effect dominates in welfare calculations, such that price matching has positive welfare implications.
The second part of the chapter tests the hypotheses with price data on the German gasoline retail market. The estimation exploits a discontinuity in the policy-affected retailers. Therefore, the analysis disentangles the competitive effects of implementing and price-consulted market participants in comparison to retailers that are not affected. As predicted, the posted average and minimum prices of one implementing retailer and its consulted competitors increase. For the other price-matching retailer, I find reduced prices that contradict the model. The last part of the chapter relates the empirics to static models and shows that the dynamic component provides previously undiscovered insights.
- Chapter 3:
The third chapter is based on joint work with Emmanuelle Auriol and Guido Friebel. It represents the subtopic of careers in this dissertation. Specifically, the chapter provides the first comprehensive data collection analysis of women’s careers in all European research institutions in the field of economics. Using a web-scraping algorithm that constantly accesses position information on institutions’ websites, we collect a novel data set on researchers in Europe. These details entail information on researchers’ gender obtained by the first name and a face recognition. Similar to survey data on U.S. institutions, we identify a leaky pipeline, as women are less likely to become professors than men are. The situation is very heterogeneous across Europe. The gap is substantially larger in Western and Southern Europe than in Central and Eastern Europe. Furthermore, we identify institutions with a higher research output and a better research-ranking having a systematically lower share of females in full professor positions as well as entry-level positions for Ph.D. graduates. Austria, Belgium, Italy, Portugal, and Spain are the drivers for this correlation. All these results are in line with the “leaky pipeline” hypothesis, in which, over the different stages of a career, the attrition of women is higher than the one of men. We show that the cohort hypothesis arguing that the lag effect between the time of Ph.D. completion and the time of promotion to a full professorship is unable to explain the current low number of females.
- Chapter 4:
The fourth and last chapter "What does Mystery Shopping do?" is based on joint work with Sidney Block, Guido Friebel, Matthias Heinz, and Nick Zubanov. It addresses an auditing practice with a yearly U.S.-turnover of 19.5 billion USD in 2016 (European Society for Opinion and Market Research, 2017: Global Market Research 2017). The term mystery represents the key aspect of the tool. During an anonymous visit, so-called mystery shoppers perform certain predefined tasks such as purchasing a product, asking questions, registering complaints, or behaving in a certain way. Following their visit, the shoppers provide detailed reports about their experiences to the evaluated firms. The chapter investigates whether the practice is suitable to determine employees’ pay. Contrary to the general understanding that firms are able to observe service quality and, in turn, can proxy for business success with mystery shopping, we do not observe mystery-shopping evaluations to correlate positively with firm performance. A decomposition of the evaluation reports indicates that mystery-shopping scores are biased and the shopper’s identity explains up to 20% of the score’s variance. Thus, the shopper’s identity has the largest impact out of all observable characteristics. With the results that mystery-shopping scores are noisy and biased, we conclude that they are not suitable for performance pay in the context of our study. In addition, we show that if the number of observations is sufficiently large, aggregated scores relate to business success. The required number of shops per evaluation period must be, however, larger by a factor between 3 and 30 per evaluated subject. Hence, cost advantages of mystery shopping diminish such that the cost benefits to customer assessments could vanish completely. The current methodology, however, may still be useful for other employee-related purposes like monitoring, which is in line with the policies of the considered firms.