Wintersemester 2019/20

30.10.2019 Sebastian Kranz, Universität Ulm
  Reconciliating Relational Contracting and Hold-up: A Model of Repeated Negotiations

Game theoretic analysis of relational contracts typically studies Pareto-optimal equilibria of infinitely repeated games. We illustrate with several examples how this equilibrium selection rules out very intuitive hold-up concerns. This becomes apparent only when moving away from the stationary environment of repeated games to more general discounted stochastic games that can have different states and can accomodate actions with long run consequences, like investments. The key problem is that Pareto-optimal equilibria, even if satisfying renegotiation-proofness, don't properly reflect plausible concerns of how today's actions affect future bargaining positions within the relationship. We propose and characterize an alternative equilibrium selection based on the explicit notion that continuation play is repeatedly newly negotiated in a relationship. We illustrate how the concept naturally combines the trade-offs from relational contracting and hold-up concerns.
13.11.2019 Guanghua Wan, Fudan University, Shanghai (China)
  Inequality and Domestic Consumption in China: A Puzzle and An Explanation

Contrary to conventional wisdom and perception, income inequality is found to be positively correlated with per capita consumption in China. This puzzle implies that fiscal policies or transfers, which aims at reducing inequality and boosting consumption, have contributed to decreased domestic demand and the slowdown of the Chinese economy. Using province panel data, we show that: (1) Income inequality is positively correlated with per capita consumption and this correlation is robust; (2) When inequality is decomposed into a between urban-rural component and a within component, only the between component is significant, implying that the puzzle may be attributable to urban-rural segregation; (3) Sub-sample estimations using urban and rural data separately show that income inequality and consumption are negatively correlated, confirming the segregation hypothesis; (4) Using urbanization rate as an indicator of urban-rural integration, inequality and urbanization (the opposite of segregation) is found to exhibit an inverted U pattern, further supporting the segregation hypothesis; (5) Finally, to investigate the mechanism underlying the relationship between segregation and consumption, we estimate the marginal propensities to consume (MPC) for rural and urban residents. It is found that urban MPC is significantly higher than rural MPC, implying a higher saving rate of rural residents despite their lower income relative to the urban residents. This lower MPC is known to be caused by the serious urban-rural segregation which limits the choice set of consumption items and increases the risks faced by rural residents, resulting in the positive relationship between income inequality and consumption. Abolishing the hukou system and all kinds of discriminations against rural residents are suggested in order to resolve this puzzle, which is imperative for preventing further slowdown of the Chinese economy and offsetting the impacts of the de-globalization wave including the Sino-US trade war.
20.11.2019 Thomas Maindl, Universität Wien (Austria)
  Optimal locations of multiple tower cranes at a construction site considering operating cost and material demand and supply

The choice of suitable tower crane configurations and locations may affect the duration and costs of construction processes considerably. The distances between the tower crane locations and material supply and demand points determine the transportation times of material and hence directly impact the operating costs. At the same time, these distances affect the required crane configurations in terms of maximum load moment and height. Here, a mixed-integer linear program (MILP) is presented that selects the optimal locations of multiple tower cranes, selects their required types and configurations, and allocates material supply points. The decision is driven by a given product demand pattern and operating costs that depend on the crane types. The specific model was developed for designing a construction site for mid-rise buildings requiring two tower cranes. While configuration-dependent crane rental costs are not part of the implemented use case, they can easily be added to the model.
04.12.2019 no seminar
11.12.2019 no seminar
18.12.2019 Thomas Gall, University of Southampton (UK)
  The value of public information in vertically differentiated markets

Generating public information about vertically differentiated products increases expected vertical differentiation and softens competition. We show that this will induce firms to overinvest (underinvest) in information generation, if the deadweight loss in the subsequent market equilibrium is high (low). Moreover, information generation by one firm has a positive externality on the other firm. It follows that coordination (e.g. via industry associations) increases information generation. When product qualities are endogenous, information generation may prevent quality degradation and thus have an additional social benefit.
08.01.2020 Denis Belomestny, Universität Duisburg-Essen
  Approximative dynamic programming

In his talk Prof. Belomestny will discuss classical results and new developments in approximative dynamic programming. In particular such topics as deep approximative dynamic programming, approximative reinforcement learning and uncertainty quantification will be discussed and illustrated by several examples. Special attention will be paid to convergence and complexity analysis of the presented algorithms.
15.01.2020 Rob van der Mei, Centrum Wiskunde & Informatica and Vrije Universiteit Amsterdam (NL)
  Saving Lives by Optimizing Emergency Response Systems

In serious life-threatening situations where every second counts, the timely presence of emergency aid (ambulance, firetrucks) can make the difference between survival or death. A promising and powerful means to reduce response time is by proactive relocation of emergency vehicles, often referred to as Dynamic Ambulance Management (DAM). Similar problem occur on maintenance systems, where a limited number of repairmen need to cope with both preventive and corrective maintenance tasks. In this talk, Prof. Dr. van der Mei will discuss new methods for proactive relocation of "resources", and show how these methods are currently operational in a number of emergency service regions in the Netherlands.
22.01.2020 Markus Weinmann, Erasmus University Rotterdam (NL)
  The Attraction Effect in Reward-Based Crowdfunding

Context effects—such as the similarity, compromise, and attraction effect—have been shown to affect choices in many situations, however, most studies use simplified (numerical) choice options and focus on "offline" situations. We test the attraction effect on a digital platform, namely, in the context of reward-based crowdfunding, using a series of experiments. Throughout these experiments, we increase realism and test various choice options, ranging from numerical to qualitative. We robustly find that even small changes to reward menus—by adding a decoy option to the choice set—can significantly influence backers’ choices and direct them to more attractive rewards. A platform's design, therefore, can have a systematic effect on users' choices, highlighting the necessity to carefully design digital platforms.
29.01.2020 Jens Brunner, Universität Augsburg
  Robust Framework for Task-Related Resident Scheduling (Authors: Sebastian Kraul, Andreas Fügener, Jens O. Brunner, Manfred Blobner)

Technological progress in health care leads to increasing complexity of the requirements in physician training. As a consequence, those programs are often not only time-related but also task-related. Task-related means that residents should perform a given number of different interventions in their program. Typically, a resident will follow a rotation across different clinical departments, where the number of performed interventions per period may be estimated. Predicting the exact number of interventions is usually not possible. Accordingly, a resident might not be able to perform all of the required interventions during the planned rotation resulting in an extension of the program. In this paper, a new model is presented that calculates the number of residents a hospital can reliably train on a strategic level. Our model also provides the corresponding training syllabi. It considers minimum requirements of both time-related stays in specific departments as well as task-related interventions that have to be performed. The robustness of the model can be set by management to handle uncertainties in interventions. A Dantzig-Wolfe decomposition is used to accelerate the solution process and a new pattern generation approach that can construct multiple patterns out of one solution is developed. The termination of the column generation algorithm is accelerated significantly by this method. The model is evaluated by a real-world case of a resident program in a German university hospital. The results demonstrate that near-optimal solutions with an average optimality gap of below five percent can be achieved within computation times of few minutes.