Reproduce “Iterative Network Pricing for Ridesharing Platforms” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chenkai Yu, Hongyao Ma
Researchers from Columbia Business School developed the Iterative Network Pricing (INP) mechanism, a model-free approach for ridesharing platforms to implement welfare-optimal origin-destination (OD) based pricing without explicitly estimating demand models. Simulations with Chicago data show INP achieves up to 99.8% of optimal social welfare and remains robust under non-stationary market conditions, including the COVID-19 pandemic.
Key insights
- Welfare-optimal outcomes in ridesharing correspond to a competitive equilibrium characterized by a uniform marginal value of a driver's time and additive OD-based price adjustments that balance driver values across locations.
- The welfare suboptimality of any origin-based market-clearing outcome can be precisely quantified using only observable market data, specifically the differences between origin-based multipliers.
- A Newton-like update direction for OD-based adjustments, computable solely from observable data (trip durations and local price sensitivities), can iteratively drive origin-based multipliers toward equality, thereby achieving welfare-optimal pricing.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.