For over a century, introductory economics textbooks have taught the same idea regarding the economics of labor: as wages rise, workers supply more labor. This idea—represented by the upward-sloping labor supply curve—has been foundational to how we think about jobs for the last few decades. But in the algorithmic labor markets of the modern gig economy, that curve is being entirely rewritten, especially as seen in Uber’s labor market
Uber has created a novel kind of labor market; no longer governed by static prices and a consistent wage, the new system is driven by real-time data and carefully-engineered behavioral nudges. The system is no longer a question of how much Uber pays, but rather how Uber motivates—even apart from wages.
The Classical Labor Supply Curve
The classical labor supply curve follows intuitive logic: more people are willing to work at higher wages. If a worker values their time at $20 an hour, they’ll only accept a job that pays more than that. As wages rise, more people are drawn into the labor market, and existing workers are willing to work even longer hours, leading the curve to slope upward.
Critically, this model assumes that workers have stable preferences and full access to market information. But the gig economy—best represented with Uber—has introduced a completely new variable: the algorithm. Rather than merely reacting to the supply and demand of the gig economy, algorithms shape it.
What is the Impact of this Algorithm?
As epitomized by Uber, the gig economy eschews the way a traditional employer operates, most notably by not offering fixed wages. Uber specifically provides a digital interface where drivers can log on and earn based on fluctuating rates, using this interface—and, perhaps even more importantly, the data that it collects—to shape worker behavior.
The company uses in-app tools and features to subtly steer drivers to supply more rides without directly raising wages. These benefits include quest benefits (extra cash if a driver hits a certain number of rides), streak incentives (where drivers are rewarded for staying online and accepting consecutive tips), and real-time reminders and push notifications (that highlight high demand areas or notify drivers that they are just one ride away from unlocking a bonus or reaching a daily goal). These benefits aren’t wage increases in the traditional incentives; rather, they’re behavioral incentives that push drivers to hit certain goals and thus increase their incentive to work longer.
The Behavioral Economics of Gig Work
What Uber has discovered is that drivers don’t behave like the rational actors of Econ 101. Rather, they’re more often influenced by psychological biases—best described with behavioral economics. For example, drivers often experience the anchoring effect, where Uber tells drivers how much they could earn if they keep driving, setting mental benchmarks that shape their decisions. They also face loss aversion: drivers don’t want to miss out on a bonus, so they keep working—not because the wage has increased, but because they fear losing a reward. Beyond that, they are victims of the sunk cost fallacy: a driver who’s been online for 5 hours might keep going, even when tired or underpaid, because they’ve already “put in too much to stop now.”
As a whole, these insights allow Uber to generate labor supply without changing the underlying wage structure. In many of these cases, the hourly rate may actually fall—but drivers keep working because of the design of the app.
A New Curve for A New Era
The gig economy doesn’t just change how people work; instead, it changes the logic behind work itself. Uber’s algorithmic labor market has revealed that the traditional model (where higher wages attract more workers) is no longer always reliable. Uber can stimulate labor supply with carefully-designed incentives, regardless of whether base pay goes up. The labor supply curve isn’t yet dead—but it’s no longer just a line on a graph.
*Image source: https://blog.afi.io/blog/using-route-optimization-to-build-a-ride-share-dispatch-algorithm/