Ride-Hailing Discrimination Metrics from Seattle and Boston Field Experiments

A dataset from the National Bureau of Economic Research by Christopher R. Knittel on discrimination in transportation network companies. It contains results from nearly 1,500 controlled ride requests in Seattle, WA and Boston, MA, recording key performance metrics. The study observed patterns of discrimination based on passenger race and gender.

Use Cases

• Train fairness-aware models for ride-hailing platforms based on observed cancellation and wait-time patterns.
• Analyze the impact of passenger demographics on service quality metrics like wait times and trip cancellations.
• Study the efficacy of anonymization interventions (e.g., removing names) on mitigating discrimination in on-demand services.

Strengths

• Data from nearly 1,500 controlled ride requests provides a substantial experimental sample.
• Results include specific, quantified discrimination patterns, such as a 35% increase in wait times and cancellation rates more than twice as frequent.
• Field experiments were conducted in two distinct U.S. cities (Seattle and Boston), allowing for geographic comparison.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• Data may reflect geographic and temporal bias inherent to the specific field experiments.

Provenance

Source

National Bureau of Economic Research

Collection Method

Field experiment where passengers were sent to hail rides on controlled routes.

Time Range

null

Freshness

Last update date is unknown; freshness unverified.

Geography

Seattle, Washington and Boston, Massachusetts, USA