When slowing down pays off: Bar-Ilan University physicists reveal surprising insights from taxi drivers

Taxi drivers who slow down when searching for passengers are not only more efficient but also earn more, according to a new study published in PNAS. The research analyzed over 2.3 billion GPS data points collected from 40,000 taxi drivers across three Chinese cities to understand which search strategies produce better outcomes, specifically, how drivers can locate passengers faster and increase profitability. The study was led by internationally-renowned physicist Prof. Shlomo Havlin from the Department of Physics and Dr. Orr Levy from the Faculty of Engineering at Bar-Ilan University, in collaboration with researchers Qiuyue Li and Daqing Li in China.

The findings revealed that drivers who “search slowly,” moving at lower speeds and making more frequent, shorter turns, achieve better performance. This result contradicts the common intuition that faster, long-range searching is more effective. Only 10% of drivers adopt this slow-search strategy, yet they earn nearly 20% more than average.

“Search efficiency is a relatively stable personal trait and does not depend on the specific day or city area,” explains Prof. Havlin. “We found that efficient drivers make more consistent turns during their search, likely due to heightened environmental awareness, and the slower they drive, the higher their chances of finding passengers.”

While the movement patterns of animals searching for food have been extensively studied in recent decades, human search behaviors have received far less attention despite being central to many aspects of life. Although most people no longer hunt or forage in nature, active search remains a core part of modern daily life. A key finding of the study is that efficient drivers’ movement patterns resemble those observed among animals searching for food. This pattern, characterized by many short movements and fewer long ones, is known to be optimal in nature. In a bustling urban environment, moving too fast can actually cause drivers to miss potential passengers.

“Even as technology evolves, with or without ride-hailing apps, efficient drivers tend to behave the same way,” adds Dr. Levy. “This study demonstrates how principles from physics and biology can help explain everyday human behavior. It also suggests that slower searching is generally more effective, because when we move too quickly, we’re more likely to overlook our target and end up spending more time finding it.”