An Undergraduate Course in the Intersection of Computer Science and Economics

Vincent Conitzer

TL;DR

This paper introduces a novel undergraduate course, "Computational Microeconomics," designed to bridge the gap between computer science and economics. It leverages linear and integer programming to teach students how to model and solve complex economic problems, making advanced topics accessible at the undergraduate level.

What Problem Does This Paper Address?

The intersection of computer science (particularly AI and theory) and economics has seen significant research growth, yet this material has largely remained confined to graduate-level courses. This creates a barrier for undergraduates who could benefit from understanding how economic phenomena arise in interconnected systems and how computational techniques can design better economic systems. The paper addresses the challenge of making these advanced topics accessible to undergraduates with diverse backgrounds, including those with limited programming experience.

Methods

The core pedagogical approach of the course is the use of linear programming (LP) and mixed integer linear programming (MILP). These techniques are presented as a powerful and unifying framework for modeling a wide range of problems in computational microeconomics. The course utilizes the GNU Linear Programming Kit (GLPK) and its associated modeling language, MathProg, which allows students to define problems abstractly before instantiating them with specific data. This approach is designed to be accessible to students with little or no prior programming experience while still challenging those who do. The course also incorporates a brief overview of theoretical computer science concepts like algorithm analysis and complexity.

Datasets

Not discussed in the paper. The paper focuses on the pedagogical approach and course structure, using example problems rather than specific datasets for evaluation.

Key Insights and Ideas

  • Unifying Power of LP/MILP: The paper highlights how LP and MILP can serve as a common language and solution framework for diverse problems in computational microeconomics, from combinatorial auctions to game theory solution concepts.
  • Bridging the CS-Economics Divide: The course successfully integrates concepts from both fields, demonstrating how CS techniques can solve economic problems and how economic thinking can inform CS system design.
  • Accessibility through Modeling: By focusing on the MathProg language, the course makes complex optimization problems approachable for students without extensive programming backgrounds, shifting the emphasis to problem formulation.
  • Front-Loading Programming: The course structure, with programming assignments early on, helps students solidify their understanding of LP/MILP modeling before tackling more theoretical or written assignments.
  • Strategic Thinking in Economics: The course emphasizes how game theory provides the foundation for understanding strategic behavior in economic mechanisms, which is crucial for mechanism design.

Key Results

The paper presents course evaluation data that generally supports the effectiveness and appropriateness of the course for undergraduates.

  • Overall Quality: Average scores for "Overall: The quality of this course" were 4.8 for the "topics" version and 4.2 for the regular course (departmental averages were 4.0 and 4.1, respectively).
  • Intellectual Stimulation: Scores for "Characteristics: Intellectual stimulation" were high, at 4.7 and 4.4 (departmental averages 3.9 and 4.1).
  • Difficulty: The subject matter was rated as moderately difficult, with scores of 3.8 and 3.4 (departmental averages 3.7 and 3.8).
  • Student Feedback: While the regular course saw a slight drop in scores compared to the "topics" version, student comments indicated a desire for more focus on mechanism design and applications, suggesting a strong interest in the material.

Limitations and Future Work

  • Limited Programming Experience: While the MathProg approach aims for accessibility, students with prior programming experience still have an advantage. The paper notes that the course is not expected to create a large number of new CS majors but can attract mathematically talented students and potentially inspire students from other fields to pursue CS.
  • Scope of "Topics" vs. Regular Course: In the second iteration (regular course), there was insufficient time to cover automated mechanism design, indicating that the breadth of topics might need careful management.
  • Future Work: The author hopes the course will inspire similar undergraduate offerings that apply computational thinking to other disciplines. The paper also suggests that the course's success could guide the design of new undergraduate courses at the intersection of CS and other fields.

Summary updated

Abstract

In recent years, major research advances have taken place in the intersection of computer science and economics, but this material has so far been taught primarily at the graduate level. This paper describes a novel semester-long undergraduate-level course in the intersection of computer science and economics at Duke University, titled “CPS 173: Computational Microeconomics.”
BibTeX
@inproceedings{conitzer2012undergraduate,
  author = {Vincent Conitzer},
  title = {{An Undergraduate Course in the Intersection of Computer Science and Economics}},
  booktitle = {EAAI},
  year = {2012},
  doi = {10.1609/aaai.v26i3.18951},
}

Loading related content...

Figure