UBER’S 70,000 UK DRIVERS ARE NOW WORKERS BUT THEY WANT MORE — WHAT WOULD THIS MEAN FOR UBER’S BUSINESS MODEL?

I met James Farrar at the 2018 Reshaping Work conference in Amsterdam. Well-spoken, humble and daring, he was flown from London to share his story regarding the landmark lawsuit he brought against Uber in the UK. I remember him saying that he picked a lengthy fight that will be very hard, if not impossible to win — it all seemed way too far-fetched, until they actually started winning…it was a game changer. James felt responsibility to stand up for himself and other drivers and was determined to see it through despite the odds.

Five years after launching the lawsuit, James Farrar, together with his ‘partner in crime’ Yaseen Aslam, helped secure social protection benefits for 70,000 Uber drivers in the UK. They have shown, and rightly so, that we can have innovative platforms like Uber that bring much consumer value and service efficiency, in combination with flexibility for drivers and adequate social entitlements.

While James and Yaseen, as union representatives, see Uber’s move to categorize all its UK drivers as workers (though the Supreme Court’s decision only applied to drivers involved in the court case) as a move in the right direction, they believe the company is not fully complying to the court’s ruling.

Following the UK’s Supreme Court ruling, Uber has promised:

· Additional payment of 12,07% of earnings towards holiday pay

· A guarantee for at least minimum wage earnings after the expenses

· A pension scheme for all eligible drivers

James Farrar and Yaseen Aslam say this is good, but not good enough. They want to see a minimum wage guarantee from the moment a driver logs onto the app to the moment of logging off. In addition, they don’t trust the company to calculate operating costs and believe this should be based on an agreement. So I wonder, how would their demands play out in this new economy and what would it mean for Uber’s business model?

UBER’S BUSINESS MODEL IN A NUT-SHELL

Uber is a ride-hailing company providing on-demand transportation services. Their added value compared to traditional taxis lays in their sophisticated technology that makes efficient matchings between drivers and passengers. Their algorithms have lowered the search time for passengers who can now get a ride with a click of a button, as well as reduced the ‘idle’ or waiting time for drivers, who can swiftly be matched to a passenger in their geographic proximity. Uber’s monitoring and GPS systems also improved passenger experience, with drivers found to take a detour less often than traditional taxi drivers, further enhancing consumer experience.

Uber is also able to adjust prices in real time based on the available supply and the level of demand. Thus, passengers may experience lower fares in the times of low demand and higher fares during the times of high demand, relative to the available supply. The so-called ‘surge pricing’ is a mechanism to increase supply at any given point in time; when the prices ‘surge’, drivers move to the areas of higher demand or decide to log onto the app, motivated by the possibility of higher pay-offs, enabling the company to meet the raising demand. Research has shown that surge prices also increase consumer welfare.

In order to satisfy the demand for rides at any given point in time, Uber has to have enough supply, that is — drivers — to meet that demand. The matter is complicated by the fact that Uber drivers are mostly self-employed and have discretion over when and where to work. This means that Uber can’t know how many drivers may log in on the app the next day, though it may be able to make a good prediction. To reduce this uncertainty inherent in its business model, Uber needs to mobilize many more drivers than it actually needs. When circumstances are different so that the demand exceeds supply, the surge pricing serves as a mechanism to motivate drivers to turn on their apps. Thus, the supply of drivers and the effort they are willing to exert — stay logged in for longer when the demand is high — must be highly elastic for Uber’s model to work. Research on TaskRabbit showed that this is indeed the case — workers respond to the raising demand by putting in more hours.

Drivers generally value flexibility that comes with the job; it grants them a better work-life balance and more autonomy over their schedule. For those that use Uber to supplement their existing income, not having to commit to shifts can be particularly advantageous. On the other hand, the situation is grimmer for those that depend on ride-hailing companies to make their ends meet. Income insecurity that comes with it, insufficient compensation for the operating costs, and lack of social protection benefits(which UK drivers will now enjoy to an extent) are some of the commonly reported disadvantages.

Considering Uber’s current business model, what would it take to compensate drivers for the waiting time and accurately estimate their operating costs?

THREE SCENARIOS SHOWING HOW AGILE UBER’S BUSINESS MODEL CAN BE

Uber’s response to the Supreme Court’s ruling has already shown us how quick and efficient the company can be at rolling out changes — this is one of the reasons we often value privately-managed solutions over the public ones. But can their digital savviness and technological superiority be used to make the pie bigger for all?

Firstly, unions argue that drivers ought to be paid for the entire time of being logged onto the app — from the moment they login to the moment they logoff. To see what this would mean for Uber’s business model, we need to understand how they earn their money.

Uber earns money from every ride given through its app. It reportedly takes 25% commission, which is where the company’s revenue comes from. The new change, specific to the UK law which has a ‘worker’ or the so-called ‘third category status’ means that Uber is required to pay drivers a minimum wage. This is the earnings floor; in theory, drivers can earn more. For instance, airport rides are usually lucrative; shall a driver have couple of those in an hour, he or she can earn significantly more than a minimum wage. If a driver however completes only few short rides that don’t add up to a minimum wage, Uber will top it up.

The problem arises with the waiting time. After completing a ride, a driver may need to wait a while before getting another ride, especially in the times of low demand, which has been the case during the pandemic. Accounting for the waiting time, they could end up earning much less than the minimum wage.

Secondly, unions argue that they don’t trust ride-hailing companies to calculate their operating expenses. Uber, for instance, commits to paying the minimum wage after accounting for the expenses (vehicle and fuel costs), which it bases on the government-determined fuel and vehicle expenses.

Below I discuss three possible scenarios that would respond to the unions’ demands and the resulting tradeoffs.

Scenario 1: No reimbursement for waiting time but access to all the necessary information

At the moment, drivers face insecurity as they cannot predict what the demand will be at any given point in time and how many other drivers may login, escalating internal competition among them. Uber provides various tools to combat that. One of those is the heat map, which shows different levels of demand in a particular area. Uber also rewards those drivers that are active during a morning rush hour, for instance, with bonus points, which entitles them to various benefits within ‘Uber Pro’ loyalty scheme. Recently, Uber also launched an ‘Earnings Estimator tool’, currently only available in France, giving drivers better insights into expected earnings. In France, Switzerland and the UK drivers are also provided with information on the fare, distance, and destination before accepting a trip. Informing drivers about their past behavior, such as rating per ride, was found to improve drivers’ future behavior, particularly for low-performing drivers.

However, the fluctuations in demand introduce uncertainty, which can for some drivers determine whether they can get by that month and pay their rents or not. To help drivers make a better choice on whether or not to work, Uber would need to ‘up’ their intelligence tool and provide real-time supply and demand information. This would allow drivers to take on ‘gigs’ with other platforms that may enjoy higher demand at that time, or take offline jobs.

In all the markets, drivers should have information on the destination, exact distance from their current location, fare charged to the customer and real-time supply and demand statistics with projected waiting time for a ride. This would allow drivers to minimize their waiting time. Uber should also provide training to help drivers navigate these tools and make adequate entrepreneurial choices.

This scenario would probably increase uncertainty for Uber as supply is likely to more closely resemble the demand, leading to the need to induce surge prices more often in order to motivate additional drivers to join. Further analysis is needed to investigate actual effects of this scenario.

Scenario 2: Reimbursement for the waiting time but less flexibility for drivers

If Uber is to pay the drivers for the waiting time, it needs to make some adjustments to its business model.

Since Uber also does not know what the demand will be at any given time, though it can make good predictions based on the passengers’ past behavior, it would need to introduce ‘shifts’. This is comparable to a number of industries, such as hospitality industry. As Fairwork rightly argues, waiters are not paid only when serving tables but during their entire shift. A restaurant, however, doesn’t make any revenue while ‘waiting for guests’ but can afford paying waiters’ wages regardless because during the times of higher demand (e.g. lunch time, dinner time) it earns more, which balances out. Waiters’ earnings also don’t change in real time as it is the case with Uber; they are fixed regardless of how busy the bar is, though waiters can collect more tips during busy times. Also worth noting is that waiters can’t leave halfway through their shifts, which Uber drivers under the current business model can.

Unlike restaurants, ride-hailing companies use market coordination mechanisms to dynamically adjust prices. This means that drivers earn less when the demand is low and more when the demand is high relative to the available supply. To compensate drivers for the waiting time, ride-hailing companies would need to introduce (1) app exclusivity (preventing drivers from multi-homing — working through the other apps simultaneously), (2) put a cap on the overall number of drivers allowed on the platform (to better optimize supply and demand), and have (3) penalty for denying rides (this would be justified since drivers would be paid to be available — so not accepting a ride request would defy that purpose and open up shirking possibilities).

App exclusivity would be crucial under this scenario as it is well known that drivers multi-home. Detecting drivers’ multi-homing would certainly be a challenge as they could be using multiple devices, which would make it hard to determine which company would be responsible for paying for the waiting time. It is also possible that drivers would be less prone to do so if they are paid for the waiting time by their chosen platform.

Drivers would be paid a fixed hourly wage and work by shifts, which would ensure having enough supply of drivers to meet the anticipated demand at any given point in time. There could, however, be a part of the workforce that can still vary and be allowed to join during the times of high demand, but that is debatable (the company may also introduce more shifts during the busy periods).

A case-scenario from New York City paints a picture of how this would look like in practice and what the tradeoffs might be. In August 2018, the New York City Council and Taxi & Limousine Commission (TLC) introduced a cap on the number of ride-hail vehicles as well as a minimum wage with a goal improve drivers’ earnings. How did the ride-hailing companies respond? Lyft introduced a ‘priority to drive’ system, limiting the number of drivers that can be on the road at any given time based on the demand. Thus, in the times of low demand, drivers need to drive to busier areas or stay offline until the demand picks up. To be on the ‘priority to drive’ list, drivers need to have above 90 percent acceptance rates and complete 100 rides in 30 days. Similarly, Uber introduced the quota system, allowing drivers to go online based on certain parameters such as ratings and the number of completed rides. Reportedly, this led to some drivers sleeping in their cars in order to stay ‘alert’ and move to the higher tier system, giving them more flexibility.

Thus under this scenario we can be certain that the total number of licensed drivers would be reduced; those that enter the system would likely earn more but would need to commit to pre-determined schedules. Ride-hailing companies would also have more incentive to treat those drivers better as the marginal cost for recruiting each new driver would be higher under this scenario. We are also likely to see more ‘professional’ drivers or those using Uber to earn full-time income and less ‘occasional’ drivers. This would thus certainly limit the opportunities for individuals that drive for Uber occasionally or to earn some extra cash.

For Uber, this does not necessarily need to result in a lower revenue but could result in reclassification in many European jurisdictions as they may be seen as employers in the eyes of policy makers due to the additional control they would need to exert over the drivers. As with the previous scenario, more research is needed to analyze the actual effects that we may see under this scenario.

Scenario 3: Allow drivers a partial control over price-setting

At the moment, most platforms either have standard, pre-determine prices (e.g. Handy) or pass the control over price-setting to their service providers (e.g. Upwork, TaskRabbit). Generally, it makes sense to do the latter when service providers have superior market-based knowledge and when the demands of customers are heterogenous. For instance, not all Upwork’s graphic designers offer the same level of quality. Some customers may just want a basic graphic design, while others may require a highly-skilled professional and would thus be willing to pay a premium price for it. Since graphic designers can assess the time needed for a specific job much better than the platform and price the service according to the quality they can offer, it makes sense to pass the control over prices to them.

Uber, on the other hand, has a relatively standardized service. Most people just want to get from point A to point B and do not necessarily care which car arrives, as long as it arrives. Uber is also able to meet the demands of customers that are willing to pay more for a luxury car via its Uber Black service, offering a certain level of service differentiation. Furthermore, it collects vast amounts of data, allowing it to gain market-based knowledge on the aggregate level. Even if drivers would have access to their data, they would only be able to analyze their own behavior — Uber on the other hand can aggregate data giving it a bird’s eye view. Customers also want to have certainty in the prices they are paying — needing to pick among different fares for the same destination would increase customer search costs and lower the overall customer experience. Thus, from what we know so far from the scholarly work, in the case of Uber, both drivers and customers would be better off with Uber retaining the control over prices.

However, recent research suggests that giving a partial control over prices may be a viable solution when there is some aspect of unobserved costs incurred by the workers. In the case of Uber, this unobserved cost is operating cost that drivers incur — vehicle and fuel costs, which vary per driver. At the moment, Uber calculates drivers’ operating costs based on the UK HMRC Guidance, calculated on averages, thus it doesn’t the variance into account. How would this partial control over price-setting work?

Uber would retain the control over price-setting, but drivers would have a margin by which they can increase prices and optimize their vehicle utilization. Such calculation would certainly not be simple on the side of the drivers but down the road this would allow Uber to collect better information on the private operating costs incurred by the drivers, which could lead to a better centrally-managed price-setting in the future.

While this does not necessarily solve the waiting time issue, it gives drivers more discretion and certainly better compensation, which they called for. Thinking along these lines — relinquishing a certain level of control may give other creative solutions for the issue pertaining to the waiting time.

The pandemic has also certainly changed the way we experience the reality. Uncertainty in the earnings and having any work at all is lingering behind and many would, at this point in time, prefer stability over flexibility. On the other hand, we are likely to have few harsh years ahead with raising unemployment and inflation. Balancing all this out and increasing labor market opportunities will require a creative act on the side of companies and policy makers.

Overall, challenging the business models of platforms and companies in general reflects a healthy and forward-looking society and it sets a precedent for better work and working conditions for future generations. I hope the three scenarios I painted in this article contribute to this end and inspire other scholars and citizens alike to engage in collective thinking in our effort to create more inclusive societies and socially responsible organizations.