Why it’s so difficult to get a cab

How often have you needed a taxi, but have not been able to find one?

Singapore has one of the highest cab densities in the world with 5,225 taxis per 1 million inhabitants, compared to 1,522 in New York and 3,285 in London.

Yet taxis in Singapore can be hard to come by, especially during peak hours and when cabbies change shifts.

This is a long-running issue, with the Land Transport Authority (LTA) introducing a taxi availability framework in 2013 to ensure that more cabs ply the roads, especially during peak periods.

The latest official data show that the situation has improved. The percentage of taxis on the roads during peak hours increased from 82 per cent in 2012 to 91 per cent in the first nine months of last year. This means about 2,000 more cabs are available for hire during peak hours.

Over the same period, the percentage of taxis driving at least 250km daily has also increased, from 75 per cent to 81 per cent.

But interestingly, the daily taxi utilisation rate, or the proportion of total taxi mileage under hire, has increased only slightly, from 65 per cent in 2013 to 66 per cent in the first nine months of last year.

So there are more cabs on the road and driving longer distances, but they are not exactly ferrying proportionately many more passengers. This is the crux of the issue, leading to perennial problems of a mismatch in taxi supply and passenger demand. Why aren’t cabbies improving their utilisation rate?

It is unlikely because of weaker demand, because average daily ridership has grown steadily from 2010 to 2014 to reach 1.02 million trips a day.

Perhaps research started in 2010 could offer some insights into the behaviour and attitudes of taxi drivers, which are still relevant today. It might explain why, even after the emergence of competition in 2013 from car sharing services and taxi-booking companies Uber and Grab, the difficulty of hailing a cab during peak hours still persists.

To better understand cabbies’ driving habits and their motivations, the National University of Singapore (NUS) Business School studied the minute-by-minute trip history of over 15,000 taxis in a month using the data furnished by the GPS satellite-based tracking and dispatching system installed in all Singapore cabs.

The study found that in a given 24-hour period, a typical cab has a passenger on board for only about seven hours. The driver spends about five hours looking for passengers, another four hours on breaks, and is offline for the remaining eight hours. The random sample included both solo drivers as well as hirers with relief drivers.

A typical shift sees cabbies having a passenger on board for slightly over 50 per cent of the time. The average shift length for a driver is 10 hours, but there are cabbies who clock less time and others who stretch their workday to almost 17 hours.

The study also looked at when a cab driver would decide to stop work for the day — after all, that is one of the key factors for cab shortages.

Tracking when cab drivers stopped collecting revenue, which could mean that their targets for daily income or driving hours were met, the study found that most stopped work even though they could go on for several more hours. This is regardless of the weather, day of the week and hour of the day.

This is not surprising, as most of us would take it easy once we have reached our target.

But in the case of cabbies, doing so means a disruption to the availability of cabs, especially near the end of the shift when their targets are likely to have been met.

What is surprising is the subsequent effect the NUS researchers observed in the two days after cab drivers exceed or do not reach their income targets.

Interestingly, their earnings went back to their usual daily level in each of the following two days. Earning less the day before did not motivate them to stay longer on the road to augment the earnings shortfall. Neither did earning more the previous day prompt them to drive less.

This pattern is consistent regardless of the day of the week, with drivers seemingly “rebooting” on a daily basis.

This does not augur well for their personal household budgeting, nor does it reflect a strong work ethic or attentiveness to customer service.

That cab drivers set daily income targets that are somewhat invariant suggests how much they work is independent of demand.

They do not appear to consider issues such as inclement weather that usually results in higher demand for taxis.

Some have argued that it is because of cabbies’ reluctance to adapt that have seen Grab and Uber flourish here. Indeed, if cab drivers want to compete against these services, whose drivers appear to be hungrier for business and more willing to be customer oriented, cab drivers must change their mindsets.

While the LTA has set standards to ensure taxi availability during peak periods, the nub of the matter is how to encourage volitional change among cabbies towards better service provision.

Taxi companies can assist in inculcating in their drivers such customer orientation.

They can also help drivers with financial planning, for example, by helping drivers track earnings on a weekly or monthly basis, perhaps with an income trajectory, so that drivers learn to see beyond the daily target.

Other possible initiatives include getting cab drivers to sign up for a recall during unexpected peak times to boost their takings while responding to a spike in demand. Encouraging cabbies to stagger the shift changing times could also increase the supply of cabs on the road.

Short of a mindset change, cab drivers and their operators will be in for a rough ride with the emergence of car-sharing services. In fact, if they do not learn anything from the competition, which has been shown to be more innovative, efficient and cheaper, they could soon be out of passengers and business.

About the Writers:

The writers are from National University of Singapore (NUS) where Sumit Agarwal is Visiting Professor at the Business School, Diao Mi and Sing Tien Foo are Assistant Professor and Dean’s Chair Associate Professor respectively in Real Estate, and Jessica Pan is Assistant Professor in Economics.