Does surge pricing really help ride-sharing drivers earn more?
- In a study of drivers in Guangzhou and Shenzhen, pricing based on supply and demand benefited cherry-picking part-timers who were motivated to work more – but discouraged full-timers and hurt their earnings
- To ensure a stable pool of full-time drivers, ride-hailing platforms might want to offer loyalty incentives
In surge pricing, used by Uber, Lyft and other ride-sharing platforms, fares go up when rider demand in an area exceeds driver supply. By encouraging more drivers to come online when needed, surge pricing balances supply and demand. This appears mutually beneficial. Impatient riders pay more and spend less time waiting, while drivers earn more during rush hour and ride-sharing platforms profit.
This economic argument holds true for rides within the surge zone. But the impact of surge pricing on driver earnings across hours and geographical zones is more nuanced.
Say a driver earns a high fare by picking up a rider from the surge zone, but if the destination is outside it – where prices might be lowered to encourage more drivers to go to the surge – the driver has to either take a lower fare next or return to the surge zone with an empty car. Worse, on returning to the surge zone, the driver has to compete with other gig drivers already in the zone.
To analyse the effect of surge pricing on driver earnings, we analysed transactional data provided by a ride-hailing platform in China. Our data is based on 4.4 million rides provided by 17,400 drivers from February to December 2019 in Guangzhou and Shenzhen.
For each day, our data includes each ride’s start and end times, travel distance and price, as well as how long each driver is on duty with the platform app turned on. In 2019, this platform was paying drivers 80 per cent of the fare with no other incentive, enabling us to determine each driver’s earning for each ride, and the daily and weekly earnings.
Uber and Lyft, on the other hand, have migrated to a more complex compensation system. For example, an Uber driver’s compensation depends on factors such as base pay, trip duration and distance, surge pricing, and incentives such as reaching a certain number of rides within a time frame. Because of this, it would be technically challenging to isolate the true effect of surge pricing in the United States.
Our data set is unique for two reasons. First, Guangzhou and Shenzhen are comparable cities with a similar gross domestic product, population size and transport infrastructure. Second, the platform introduced surge pricing first in Guangzhou on October 11, and then in Shenzhen on December 11 in 2019. This two-month gap made for an ideal experimental setting.
From October 11 to December 10, drivers in Guangzhou experiencing surge pricing (the so-called treatment group) can be compared with drivers in Shenzhen with no surge pricing (the so-called placebo group). By looking at the number of trips and earnings between the two groups, we can determine the causal effect of surge pricing on driver earnings.
The first result was that the Guangzhou drivers experiencing surge pricing earned 13.7 per cent more per week on average than the Shenzhen drivers with no surge pricing. This appears to confirm the belief that surge pricing helps drivers earn more. But this does not account for how the Guangzhou drivers had changed their behaviour in response to surge pricing.
02:05
China’s ride-hailing drivers install plastic shields to prevent coronavirus infection
By comparing driver behaviour and earnings, we obtained our second result. Relative to the Shenzhen drivers, the Guangzhou drivers tended to switch their working hours to concentrate on rush hour in the hope of getting more surge-priced rides. But, as more drivers turned on the app during rush hour, competition intensified. As a result, they actually got 8 per cent fewer rides and earned 7 per cent less daily compared with the Shenzhen drivers – too many were competing for surge-priced rides during rush hour.
But there still remains the intriguing finding that Guangzhou drivers were, on average, earning less daily yet seeing a boost in their weekly takings. This brings us to our third result, which found that, to compensate for the reduction in daily earnings, the Guangzhou drivers had increased their working days by 12 per cent a week.
So far, these results represent the average impact of surge pricing across all drivers in the Guangzhou group. But the impact of surge pricing can be different for part-time and full-time drivers, especially when the full-timers have to compete with cherry-picking part-timers during rush hour.
So we classified drivers in each city into two groups: 74 per cent were part-timers who worked 1.4 days a week on average and 26 per cent were full-timers who worked more than five days a week on average.
We unearthed an interesting result. While surge pricing increased the size of the pie, with total revenue increasing by 13.7 per cent, the size of the driver pool also increased, by 20 per cent. As competition intensified, each driver received a smaller piece of the pie in terms of earnings and surge-price rides.
More important, we find that as the pie slices become smaller, surge pricing had the effect of discouraging full-timers, who worked 2 per cent fewer hours and earned 7.8 per cent less per week. But the part-timers, who have the capacity to work more, increased their hours by 17 per cent and, as a result, earned 21.3 per cent more per week. Their increased participation compensated for the drop in full-time drivers.
Overall, surge pricing intensifies competition and does not benefit all drivers. It hurts full-timers, even though it is beneficial to part-timers who work more hours during the week. To ensure a stable pool of full-time drivers, ride-hailing platforms might want to offer incentives such as a bonus or benefits for loyal drivers.
Dr Christopher S. Tang is a professor of supply-chain management at the UCLA Anderson School of Management
