Matching & Marketplaces

Foundational theory introducing price structure vs. price level distinction for platforms.

Comprehensive survey defining two-sidedness by price structure; multi-homing analysis.

Three canonical models (monopoly, single-homing, competitive bottlenecks); workhorse for policy analysis.

Addresses startup problem and price competition; divide-and-conquer strategies.

The theoretical gold standard for platform pricing; introduces Spence distortion showing platforms internalize network effects only for marginal users. Essential for welfare analysis.

Studies sequential arrival of sides and commitment decisions; key for understanding multi-homing dynamics and platform timing strategies.

Explains how tech giants expand across adjacent platform markets through bundling; foundational for understanding winner-take-all dynamics.

Comprehensive summary of canonical search-matching theory; matching function, wage bargaining, unemployment dynamics.

Definitive textbook treatment of DMP model; the practitioner's reference for labor market search.

Empirical estimation of aggregate matching functions; bridges theory and data.

Introduces endogenous job destruction into search framework; stochastic equilibrium model.

Field experiment on Upwork (oDesk) demonstrating information frictions cause inefficient unemployment; first rigorous platform labor experiment.

Uses Upwork data to quantify how platform certification helps inexperienced workers overcome reputation gaps and find first jobs.

Amazon Mechanical Turk experiments find substantial monopsony power (labor supply elasticity ~0.1) facing individual requesters.

Upwork analysis showing workers capture surplus because demand-side search frictions reduce direct competition between workers.

Introduces deferred acceptance algorithm and stability concept; foundation of market design.

Applies mechanism design to organ allocation; created real kidney exchange programs; saves thousands of lives.

Engineering economics: redesigning NRMP; demonstrates stability matters empirically.

Shows ride-hailing prone to matching failures; explains why surge pricing is necessary.

Shows platforms can facilitate efficient market clearing through targeted recommendations; reduces wasted applications and improves welfare for both sides.

Foundational queueing model for Uber/Lyft showing dynamic pricing is robust to uncertainty; establishes theoretical foundation for surge pricing.

First rigorous empirical analysis of online reputation (eBay); documents extreme positive bias.

Field experiment proving ~8% reputation premium; gold standard methodology.

Documents incentives for fake reviews; 16% of Yelp reviews suspicious; competition drives fraud.

Airbnb experiment showing bilateral reviews cause reciprocity/retaliation bias.

eBay analysis showing reputation mechanisms suffer from reputational externalities and silent dissatisfaction bias; introduces 'Effective Percent Positive' measure.

First rigorous fake review identification comparing TripAdvisor vs. Expedia; shows independent hotels strategically post fake reviews of competitors.

Classic on voluntary disclosure in online markets; shows sellers reveal private information to reduce adverse selection in used car sales.

Uses surge pricing variation to estimate demand; $6.8B consumer surplus; RDD methodology.

Structural model of surge pricing welfare; riders gain, drivers lose; distributional analysis.

Optimal overbooking with multiple fare classes; airline revenue management workhorse.

Shows restricting visibility reduces wasted applications and improves welfare for both sides; directly applicable to Upwork, dating apps, Airbnb.

Proves equivalence of independent lotteries and penalty-waitlists; applicable to any platform managing queues for scarce resources.

First comprehensive Uber labor market analysis using administrative and survey data; foundational descriptive paper on gig work.

Estimates drivers earn more than twice the surplus they would in rigid arrangements; seminal flexibility valuation paper.

Field experiment showing rideshare drivers strongly prefer proportional commission over fixed taxi leases; finds intertemporal substitution elasticity ~1.2.

Documents 7% gender gap among Uber drivers entirely explained by experience, preferences, and driving speed—not discrimination.

Field experiment finding average worker WTP 20% of wages to avoid employer-set schedules; explains gig work appeal.

Documents alternative work growth from 10.7% to 15.8% of workforce; contextualizes gig economy scale and trends.

Foundational theory showing surge pricing benefits all stakeholders by better utilizing flexible capacity; explains when dynamic pricing is welfare-improving.

Demonstrates Q-learning algorithms autonomously learn collusive strategies; major antitrust implications for platform pricing.

First large-scale empirical analysis finding margins increased 28% when both competitors adopted pricing algorithms.

First paper documenting naive A/B testing overstates treatment effects by ~2x due to general equilibrium effects; eBay data, launched this literature.

Comprehensive treatment of experimental design including interference; essential methodology paper for platform economists.

Characterizes when to randomize on supply vs. demand side; practical guidance for platforms running experiments.

Meta-experiment on Airbnb empirically measuring interference bias magnitude; bridges theory and practice for marketplace experiments.

Shows 42% of booking inquiries rejected by hosts; platform search design critically affects transaction volume and market thickness.

Exploits Craigslist's staggered entry; landmark empirical paper on platform competition and displacement effects.

Amazon Marketplace data showing Amazon targets successful product spaces for entry; essential for self-preferencing debates.

Estimates 8-10% causal impact on hotel revenue in Austin; shows sharing economy disruption dynamics and incumbent responses.

RCT showing enhanced trust increases hiring but also disintermediation when trust is sufficiently high; platform design tradeoff.

Uses China Skype blockade as natural experiment; finds restricting communication reduces disintermediation by 18%.

Uses eBay browsing data to show narrowing consumer choice sets can paradoxically strengthen seller competition and lower prices.

Structural model quantifying welfare: consumer surplus $41/night, host surplus $26/night; benefits concentrated during peak demand.

Comprehensive framework covering search, matching, pricing, and reputation in peer platforms; essential survey paper.