Michael Sheldon

Economist × Data scientist



I am a Data Scientist at Uber Technologies, where I research the economic behavior of drivers. I work on projects regarding driver structural and realtime pricing, developing algorithms and experimental designs to optimize these separate pricing policies.

With Keith Chen of UCLA, I have studied the reaction of driver supply to unanticipated short-term variation in earnings; our current working paper discusses behavior heuristics such as "income-targeting" and the role they play in the supply of labor. 






LinkedIn Profile


Downloadable Format

Michael Sheldon



The University of Chicago

Bachelor of Arts in Economics with Honors, March 2016
GPA: 3.75/4.00
BA Thesis; Income Targeting and the Ridesharing Market
David S. Hu Award for excellence in course work and creativity in thesis


Data Scientist II

Uber Technologies, San Francisco California
June 2014- Present                       

  • Lead contributor for a system which creates, optimizes, and updates driver incentive campaigns globally among top cities. Built on iterative experimental design and a gradient descent approach to optimize incentive spend between structures, geographies, and times within a city.
  • Developed a modeled framework to understand market-level impacts of supply and demand experiments, which typical A/B experiments cannot detect. Led project to extend this framework in order to analyze the competitive impacts of network scale.
  • Key contributor to decoupled, driver surge pricing. Developed a new surge algorithm for drivers which better conforms to supply behavior and elasticities.                                                                                                       

Research Assistant to Devin Pope, Behavioral Economics

University of Chicago, Booth School of Business
February 2014-March 2016                                                                                                

  • Designed and managed a 10,000+ participant study on MTurk, analyzing task performance under different behavioral frames and incentives
  • Led a group of 10+ individuals for collecting, compiling, and cleaning race data on police shootings.
  • Provided preliminary insights, data collection and cleaning for play-by-play NFL data                                                                              

Project Development Assistant

Collective Decision Engines, Chicago Illinois
March 2014- June 2014

  • Involved with developing a commercial implementation of Quadratic Voting under Glen Weyl
  • Conducted a user-research panel to test software functionality and interpretability.
  • Developed regression testing of the software prior to commercial implementation.



#BUILTBYGIRLS, San Francisco California
February 2018- Present

  • Provided advice, guidance, and mentorship to young women looking to enter and succeed in the tech industry.
  • Shared experience and insights, provided advice for opportunities and growth, and introduced mentees to other experts in their desired industry    

Saturday University Tutor

Black Star Project, Chicago Illinois
April 2013- June 2013

  • Instructed group of at-risk students in both mathematics and reading
  • Coordinated assignments, evaluations, and curriculum with other instructors                                                                                

Chicago BounD

September 2013

  • Helped create partnerships between non-profit organizations and the University of Chicago
  • Engaged and consulted community members in addressing social issues such as LGBTQ+ homelessness, reemployment programs, and recidivism


Technically Proficient in: Python, R, Hive, Presto, SQL, STATA, LaTeX
Language: Proficient in German

Working Papers

Dynamic Pricing in a Labor Market: Surge Pricing and Flexible Work on the Uber Platform

Joint with Keith Chen

Last Edited: 12/11/2015

This paper studies how the dynamic pricing of tasks in the “gig” economy influences the supply of labor.  In this paper, we study how driver-partners on the Uber platform respond to the dynamic pricing of trips, known as “surge” pricing. In contrast to income-target findings, we find that Uber partners drive more at times when earnings are high, and flexibly adjust to drive more at high surge times. A discontinuity design confirms that these effects are causal, and that surge pricing significantly increases the supply of rides on the Uber system.

Income Targeting and the Ridesharing Market

Last Edited: 02/18/2016

This paper examines the supply elasticity of individual contractors in the ridesharing market. Contrary to the results of Camerer et. al (1997) and the "income targeting" hypothesis, this study demonstrates substantial evidence of positive labor supply elasticities. Furthermore, the paper discusses the nature of measurement error in labor markets and how its presence can severely bias results downward.