Empirical Applications of Game Theory

1 Description

This unit is about game theory and its applications to economic and social decision problems. Economists and other social scientists use game theory to understand, model, and predict people’s behavior in strategic situations. These are situations where the welfare of any person depends not only of what that the person does himself, but also on the decisions made by other people. Game theory has application in all areas of economics and many economic and other social situations require the use of tools from game theory. For example, game theory is used to analyze issues related to competition policy, auctions, matching, bargaining, status signaling, education choices, new technologies adoption, patenting, betting markets, insurance, the emergence of social norms, trade and environmental agreements, incentive systems, etc… However, economists rarely test whether game theoretic explanations for behavior actually apply in the field. This seminar therefore proposes to examine how game theory can yield testable empirical predictions and solutions to economic and public policy problems. This goal will be achieved by asking participants to develop a precise understanding of a strategic situation of their own choice. Participants will describe and model the strategic situation, and then solve the model and compare their solution with the way the game is actually played in the field.

2 Organization

1. April 17 to May 15: I will explain the aim of the course and present a few examples of applications of game theory to a range of economic situations (See section 7). We will learn how to confront empirical evidence with predictions from game theory. We will also talk about how to model original and interesting economic situations as games.
2. May 15 to June 26: Participants will have to find a topic for further empirical investigation (See examples in section 8). They will develop a research strategy to explore and understand this topic, and plan and carry out data gathering to empirically test their theoretical modeling and predictions. I will be available for individual consultations (see section 6).
3. June 26 and 27: Participants will present and discuss their findings, and receive comments from other students and from the lecturer. This presentation will count for 50% of the grade.
4. July 11: Deadline for the submission of essays. The essay will count for 50% of the grade.

3 Textbooks

1. Gibbons, R. (1992). Game theory for applied economists. Princeton University Press.
   (for students in need of catching up with game theory).
2. Holt, C.A. (2007). Markets, games and strategic behavior. Pearson.
   (covers a range of economic situations, but uses mainly experimental data).

4 Examination (6 Credits)

The grade for this course is divided into two parts, 50% based on an oral presentation and 50% based on an essay. Individual essays are 15 pages, written in either English or German.
In the presentation and essay, participants will have to

1. Explain why their topic is important and show they understand the economic situation they chose to analyze.
2. Describe the players, how they interact, and represent the situation as a game in extensive form.
3. Use game theory to get a prediction of the outcome.
4. Give an economic/social explanation and interpretation of the predicted outcome.
5. Compare predictions of the outcome with empirical and experimental observations.

The presentation and essay also require:

1. Independent literature research and data gathering.
2. Clear written presentation of the research questions and its theoretical connections.

5 Dates & Locations:

Course components	Times	Dates	Location
Seminars	Wednesdays, 10:00 – 12:0	17.04.2019 to 15.05.2019	Oec 1.164
Presentations	Wednesdays, 10:00 – 18:00	26.06.2019 and 27.06.2019	Oec 2.157
Essay deadline	11.07.2019

6 Contact information

I am available for discussion and advice-giving as you go along with defining, implementing and writing up your project. Please write me an email at a.gaudeul@gmail.com to agree a time for a meeting. My office is in room 8.124 in the Blue Tower (Mehrzweckgebäude, MZG).

7 Literature

We will discuss the following articles during the first four weeks. The articles were selected to cover a range of applications of game theory in different domains. We will focus on how the authors identified interesting economic situations that can be modeled with game theory, and how they went about expressing hypotheses, gathering data and analysing it.

1. Palacios-Huerta, I. (2003). Professionals Play Minimax. The Review of Economic Studies, 70(2), 395–415. http:// https://doi.org/10.1111/1467-937X.00249
   The paper discusses whether we are actually able to play mixed strategies, and tests this using data from professional football games about how players shoot penalty kicks.
2. Stephan, M. J., & Chenoweth, E. (2008). Why civil resistance works: The strategic logic of nonviolent conflict. International security, 33(1), 7-44. https://www.mitpressjournals.org/doi/abs/10.1162/isec.2008.33.1.7
   This paper examines theoretical reasons why non-violent political campaigns are often more successful in achieving their aims than armed campaigns. The theory is tested with data on major nonviolent and violent resistance campaigns from 1900 to 2006.
3. Lin, J. (1990). Collectivization and China’s Agricultural Crisis in 1959-1961. Journal of Political Economy, 98(6), 1228-1252. http://www.jstor.org/stable/2937756
   This paper asserts that preventing members of Chinese agricultural collectives from leaving their collective changed incentive structures and led to a breakdown in productivity. The author carefully examines other hypotheses.
4. Ashenfelter, O., & Genesove, D. (1992). Testing for Price Anomalies in Real-Estate Auctions. The American Economic Review, 82(2), 501–505. https://www.jstor.org/stable/2117452
   This paper tests for evidence of a winner’s curse in real-estate auctions. The winner’s curse is such that those who win auctions tend to pay too much for what they get. The authors test this by comparing prices obtained in auctions with prices obtained in face-to-face negotiations.
5. West, D. S., & Hohenbalken, B. V. (1984). Spatial Predation in a Canadian Retail Oligopoly. Journal of Regional Science, 24(3), 415–427. https://doi.org/10.1111/j.1467-9787.1984.tb00812.x
   While backward induction would predict that established firms cannot deter entry by new firms in their markets, there are a number of models that rationalize entry deterrence. However, empirical evidence is hard to come by. This paper considers reaction to entry in the retail market of two Canadian regions.
6. Shankar, V., & Bayus, B. L. (2003). Network effects and competition: An empirical analysis of the home video game industry. Strategic Management Journal, 24(4), 375-384. https://doi.org/10.1002/smj.296
   When there are network effects for a product, then the firm with the largest user base will have a competitive advantage and be able to maintain its dominance. The authors use data on competition between Sega and Nintendo to show that the firm with a smaller user base was able to overtake the larger one because network effects were stronger for its users.
7. Mavlanova, T., Benbunan-Fich, R., & Koufaris, M. (2012). Signaling theory and information asymmetry in online commerce. Information & Management, 49(5), 240–247. https://doi.org/10.1016/j.im.2012.05.004
   Signaling theory predicts that players in a game can credibly signal their type (quality of their product, intelligence, strength) by using costly signals. The authors examine this theory by considering whether high-quality sellers of medicine online use more costly signals of quality than low-quality sellers.
8. Rauchhaus, R. (2009). Evaluating the Nuclear Peace Hypothesis: A Quantitative Approach. Journal of Conflict Resolution, 53(2), 258–277. https://doi.org/10.1177/0022002708330387
   Mutually assured destruction is a military doctrine according to which two players with the means to annihilate each other will be less likely to initiate conflicts. The authors show that nuclear weapons indeed lower the likelihood of conflict, but only if there is parity between conflicting parties. If one side has more nuclear weapons than the other, then nuclear weapons make conflicts more likely.
9. Davies, R. B., & Vadlamannati, K. C. (2013). A race to the bottom in labor standards? An empirical investigation. Journal of Development Economics, 103, 1–14. https://doi.org/10.1016/j.jdeveco.2013.01.003
   When countries are competing to attract capital, then they may be tempted to reduce labor standards, taxation or environmental regulation. This paper tests this hypothesis in the case of labor standards and shows that it holds mainly for developing countries that already have weak standards.
10. Bandiera, O., Barankay, I., & Rasul, I. (2005). Social Preferences and the Response to Incentives: Evidence from Personnel Data. The Quarterly Journal of Economics, 120(3), 917–962. https://doi.org/10.1093/qje/120.3.917
    Principal-agent models have been used to devise contracts to motivate employees to work harder. For example, one can use tournaments, whereby pay depends on relative performance vs. other workers. The authors show however that workers internalize the effect of making more effort on the welfare of their colleagues, which means they do not work as hard as would be predicted if they were egoistic decision-makers.

8 Indicative list of possible topics

1. Competition in the ride-hailing business, for example how firms such as Über or Lyft compete to attract drivers and passengers.
2. Queuing strategies, with data from observation of queuing at supermarket counters.
3. A comparison of positioning strategies for large and small software companies, with data from observation of software review websites and online advertising.
4. Status signaling, with data about car choices in the German automotive market.
5. How do people learn social norms, and when do they choose to challenge them? With observations of systems for the social integration of immigrants in different countries.
6. Playing games with one’s future self: Can the use of behavioral commitment devices improve one’s welfare? With observations of the drop out rate in the use of fitness tracking apps.
7. “Governing the commons”: A comparison of the organization and success of different arrangements for the upkeep of WGs (Wohngemeinschaften, apartment-sharing communities) in Germany.
8. Strategies for promoting growth in online communities: the balance between censorship and defense of free-speech. With comparisons of policies at different Reddit communities.
9. How big can a conspiracy be before it is discovered? With data on collusive agreements, how long they have been maintained and how many participants were involved.

Advertisement