Daniel Chen

Problem Definition: Gig economy workers allocate their labor supply strategically, with wider consequences for the operations of firms, worker welfare, and the regulation of labor markets. Using a dataset tracking individual drivers' jobs and job search activities across all ridehailing and delivery apps, we study the drivers' strategic behavior in repositioning spatially and in multihoming between platforms.
Methodology/Results: By estimating a structural model of how drivers optimize their earnings across both platforms and locations, we find that their heterogeneous costs for engaging in strategic multihoming are large and significantly outweigh their costs for repositioning.
Managerial Implications: We evaluate regulations and gig market reforms that affect strategic multihoming and repositioning. We find that markets face a general tradeoff between raising worker earnings and positive effects on market expansion and downstream inter-platform competition: the second-order effects of competition between workers generally dominate the first-order gains in earnings that accrue to workers with more flexibility. When given the frictionless freedom to multihome, workers' hourly earnings and utilization fall by 8.1% and 6.9% respectively even as the market expands by 3.8%. Interestingly, the current equilibrium in the New York City marketplace for ridehailing and delivery gigs approximates one with no multihoming, which would reduce service volume by only 1.4% and raise drivers' hourly earnings by 3.3%. Imposing even small repositioning costs can broadly alleviate congestion.

Machine learning is a growing method for causal inference. In machine learning settings, prediction errors are a commonly overlooked problem that can bias results and lead to arbitrarily incorrect parameter estimates. We consider a two-stage model where (1) machine learning is used to predict variables of interest, and (2) these predictions are used in a regression model for causal inference. Even when the model specification is otherwise correct, traditional metrics such as p-values and first-stage model accuracy are not good signals of correct second-stage estimates when prediction error exists. We show that these problems are substantial and persist across simulations and an empirical dataset. We provide consistent corrections for the case where unbiased training data is available for the machine learning dataset.

Algorithms control pricing and match customers and workers in the gig economy. Despite their prevalence, algorithms face several critiques: they lack transparency, can be biased, and can be inefficient. We collaborate with a driver analytics company to empirically analyze these issues using a comprehensive dataset on completed trips in the Chicago ridehail market. We show that algorithmic pricing varies significantly after controlling for distance, duration, location, and time of day: customer fares have a residual standard deviation of 36% ($6.16 against an average of $17.12 per trip), while worker earnings vary 29% ($3.77 against an average of $12.86). We find that algorithms are not biased against low-income locations, as ridehail platforms offer lower prices and higher service levels relative to taxis. However, algorithms lose efficiency from two sources: inefficient routing from competition between platforms decreases worker utilization by 1%, and incentives for workers to decline trips further decrease utilization by 11%. We model workers’ strategic responses to variation in pricing and estimate counterfactuals on the effects of minimum wage and transparent pricing policies.

Online ratings are a prominent feature of e-commerce and social media platforms, but crowdsourced ratings are often unreliable signals of quality. We show that network information can significantly improve signal quality by mitigating difficulties from bots, strategic users, and heterogeneous preferences. First, using theoretical models, we illustrate several advantages of network-based ratings for converging to accurate ratings and preventing herding to undesirable outcomes. Second, using large empirical datasets from e-commerce and social media, we show that improvements from network ratings are large and practical to implement.

The emergence of e-commerce and digital operations has given firms unprecedented reach: retailers can now send real-time emails and mobile notifications to customers anywhere in the world. These direct-to-consumer promotions (e.g., coupons, emails, push alerts) can boost engagement and sales, but when overused, they can create annoyance and notification fatigue, alienate customers, increase unsubscriptions, and ultimately weaken long-term customer spending. In this paper, we first conduct a large-scale randomized field experiment at an online retailer to study this trade‐off between engagement and annoyance. We find that sending fewer emails significantly reduces unsubscription (proxy for churn) by 59% and thus increases future customer survival, but at the cost of a 5-8% decrease in short-term revenue and purchase rate. These effects are heterogeneous: reducing coupon-email frequency lowers churn, especially among customers who previously used few coupons but opened many emails, and it reduces short-run spending, particularly among higher-volume customers and those who rarely opened emails. Then, using historical data over a longer time frame, we find that unsubscribing has a large negative effect on monthly revenue, with a 36% decrease in customer spending. We then embed these results in a customer lifetime-value (CLV) optimization model. The model shows that personalizing email frequency using the field-experimental estimates can improve long-term value: a myopic policy (maximizing immediate revenue) raises average CLV by 6.8%, while a full CLV-optimization policy yields an 8.3% gain over the daily-email baseline. These gains are driven by higher retention as well as modest changes in coupon usage. Overall, our findings highlight the importance of balancing short-term revenue with customer retention: sending emails with coupons too frequently can boost immediate sales but may "burn out" customers, whereas moderately reducing email frequency can yield higher lifetime value in many settings.