A visualization of "Algorithmic Price Discrimination - Possible Welfare Outcomes of 3rd order price discrimination with noisy measure of buyers valuation".
This small project simulates the welfare implications of 3rd order price discrimination with uncertainty about the buyers' product valuations. The code implements an easy example of Cummings et al. (2020) to provide an inuition for their theoretical results. The paper and the specific example analyzed here is described in this Presentation, Section "Noisy Signal". The project was a part of a Master-level class in Industrial Organization by Prof. Stefan Buehler and Dr. Nicolas Eschenbaum at University of St. Gallen.
Cummings, Rachel Devanur, Nikhil R. and Huang, Zhiyi and Wang, Xiangning. "Algorithmic Price Discrimination". 31st Annual ACM-SIAM Symposium on Discrete Algorithms (SODA 2020), Salt Lake City, UT, USA, 5-8 January 2020.
Bergemann, Dirk and Brooks, Benjamin and Morris, Stephen. “The Limits of Price Discrimination.” The American Economic Review, vol. 105, no. 3, 2015
Erik Senn (erik.senn[at]gmx.de), Katharina Haglund, Jeremia Stalder
main.R - set input parameters, generate segmentations, plot results
functions.R - functions to get welfare outcome for given segmentation
Idea: Compute possible welfare outcomes of market segmentations with only a noisy measure of buyers valuation.
Output: Dataframe with optimal outcome for each tested segmentation, plots for welfare outcomes with and without segmentations
Current Setting: Example of Cummings et al. Simple random noise and 3 valuations / types.
Limitations: Only supports length (value set) = length (type set). Special points (red) in plots are hard-coded, might not fit for every noise level.
- Compute optimal segmentation via linear programm (Cummings et al., Theorem 2.1).
- Define other noise functions.