Taxi drivers, on the other hand, complain that there aren't enough customers. For the past few years, they have been saying that it's really hard to make ends meet, after forking out about S$90 a day for rental of the vehicle from a taxi company, the ever increasing cost of fuel not included. The result is that they find themselves having to drive for up to 12 hours to earn enough, but this then means they get so tired, they become a road hazard.

So, is there a shortage of taxis in Singapore?

Not according to the Land Transport Authority (LTA). With one registered taxi per 180 persons, it's better than the ratio in many other cities around the world, they said.

This may be a little misleading though, as a friend related to me recently what she had heard from a taxi-driver. According to him, there are fewer drivers than available taxis, since word has gone around about how difficult it is to make a living from driving.

This may be borne out by the statistics I found on the LTA website showing a falling number of taxis through the first 3 months of this year. The taxi companies may be retiring old vehicles without replacing them when they can't find enough drivers.

That notwithstanding, I suspect it isn't a macro, but a micro problem. By that, I don't meant to dismiss it as a small problem, but to suggest that it may be less a problem of overall numbers, than one of resource allocation.

Demand spikes at certain times (and in certain directions only) e.g. during the evening peak when commuters want cabs from downtown to the suburbs. Supply evaporates at other times, e.g. during the Cinderella hour.

Commuters may not have any problem getting a taxi 9 times out of 10, but on the one occasion when they can't get one, they really, really can't get one. They find themselves waiting by the side of the road or in a taxi queue for an hour, or they have to make a frustrating number of calls to the radio despatcher without success.

Hence, despite the persistent outcry, there may not, overall, be any excess demand over capacity. Thus, to propose increasing fares across the board to moderate demand would be foolish. It has been reported that taxi-drivers themselves did not like the idea when it was recently floated by the taxi companies; they were afraid that the driving-around-empty afternoons would get even emptier.

What about increasing fares during the peak periods? This may not work too, for demand may not be elastic enough. But first, we should think critically about the use of price signals.

The quick resort to higher fares to solve problems is entirely consistent with the Singapore belief that in economic management, we apply market principles unlike some other authoritarian states that prefer to rely on administrative controls.

However, this belief that we're applying market principles may be false.

For price signals to work, there must firstly be realistic alternatives to the consumer -– and if they do not yet exist, there should be no significant barriers to new market entrants who are prepared to offer new products and services. I will discuss below why I do not think there are, at present, viable alternatives to the taxi for many consumers. The absence of alternative modes of transport means that demand elasticity may not be sufficient for price signals to work. Even if price hikes are high enough to deter consumers from using taxis, there will be a political price to pay -– consumer anger.

Secondly, relying on prices to smoothen out supply and demand is meaningless without enough competition in the marketplace. In many ways, taxi fares are not really set between suppliers and consumers, but between suppliers and regulator. 

Of course, I don't mean to suggest that fares should be determined by haggling between drivers and potential customers. But if there is no dominant player in the taxi market, and no cap on the number of taxi licences available, there will likely be greater fluidity in taxi pricing, eventually adjusting in response to consumers' notion of value.

On top of that, at the present moment, in the absence of competitive pricing, fare structures tend to be determined by marking up on cost. Whenever the taxi companies want to increase their fares, they have to show the regulator cost reasons for it. For all practical purposes then, fares are thus set by reference to cost, not by reference to competitive pressures. 

(Which may be why we so seldom hear the dominant taxi company complaining of poor profits.)

In short, taxi operations in Singapore rather resemble central planning, with pricing by administrative fiat. Bottlenecks are the usual consequence of central planning; we should hardly be surprised that bottlenecks are what we have.

More specifically, the Cinderella problem is entirely a problem made by central planning.

Taxis disappear from our roads after 11 pm because drivers know they will make more money after midnight when a 50% surcharge kicks in. Consumers try to get a cab before midnight to avoid the surcharge. The result is not hard to guess.

Why does Singapore have a midnight surcharge when other cities do not? Because decades, yes decades, ago, some civil servant believed that taxi-drivers would not be motivated to work after midnight otherwise.

That was when Singapore had virtually no nightlife and demand was much too thin. But is that assumption true today? Would there be no taxis without the surcharge? Nobody has bothered to reexamine it.

As well, is 50% the appropriate surcharge? Why not 25%? Why not 15%? By introducing such a major price break-point, it should hardly surprise anyone that it disrupts the market severely. A more freely functioning market, with real competition, would have found a smoother formula through interactions between (competing) suppliers and commuters.

As mentioned above, pricing may not be a solution to the peak hour problem either, because it's basically a surge effect. If we are ever to have enough taxis to serve all the commuters going home, then we'll have far too many taxis cruising empty on the return leg or at other times.

The way to handle surge effects is to look for alternative ways to cope. We can begin by asking: why do commuters want to take a cab home, instead of buses and trains? I don't think the answer is hard to find. I can think of these likely reasons:

• The buses are packed like tinned sardines;
• The trains are packed like tinned sardines;
• The trains' and bus' routes are planned in such a way that nearly always, one has to change modes to get to one's final destination;
• Changing modes means lots of waiting time for the next bus or train to arrive.

In short, public transport is time-consuming, tiring and generally a hassle, which means it's essentially a failure of planning.

I will argue below that bus and train transport is designed for a narrowly-defined reference commuter. Any commuter who does not fit that reference profile is outside the planning framework.

We saw this some time ago when members of the public asked why we never built elevators at MRT stations, thus barring wheelchair-bound persons from using trains. The response from the LTA was that they should use taxis.

Immediately, the public grasped the Marie Antoinette nature of this reply. On the eve of the French revolution, ministers told the queen that the peasants were starving, they having no more bread. "Let them eat cake", she was reputed to have replied, showing how out of touch she was with the plight of the poor.

The tunnel vision is the failure to see that not everyone can afford to take taxis everyday and that it was negligent of the LTA to plan only for a narrowly-defined reference commuter.

Another tunnel vision effect is for the planners not to take fully into account waiting times when changing mode of transport. It's no use telling people that trains and buses travel smoothly and quite speedily, if they take forever to arrive.

What I have noticed is that in terms of route management, there is a fondness for the hub-and-spoke model. The idea is to integrate buses and trains. But as anyone who has ever faced a 3-hour layover at an airport waiting to catch a connecting flight can attest, the hub-and-spoke model has a serious drawback -– the time wasted in waiting for the connection.

For people who have had a long day at the office, perhaps working overtime as well, hoping to rush home to be in time to see the kids before bedtime, a 75-minute, multi-leg commute is not something to look forward to. 

Why is our public transport so much hub and spoke? Because the 2 bus operators are also train operators in their respective zones. For them, it may be the most efficient way of doing things.

This is compounded by the tendency to assign parts of Singapore to these 2 companies as their near-exclusive zones. As a consequence, they rarely compete against each other route for route. What we have is not a competitive market but a partitioned duopoly.

A more open market (with more than 2 bus companies) should see new point-to-point routes develop. As a result, more commuters may enjoy having a direct bus from work to home instead of having no choice other than a tedious multi-mode journey. If it seems too tedious, there will be a strong temptation to opt for a taxi instead, thus adding to the surge.

The other big grouse is the way buses and trains are jampacked after 6 pm. Our transport planners somehow think that capacity is sufficient when in the eyes of many consumers it clearly isn't.

Another packed train, even at around 10.30 pm

Their planning may also be based on a flawed assumption: that the typical commuter is a single person with only a newspaper in hand. If you're carrying a briefcase, plus 2 bags of groceries and dragging a child along, our public transport serves you very badly. Squeezing into a bus or train, and then strap-hanging all the way home with a screaming child is not something you want to do. So you wait for a taxi.

Solving the taxi problem must involve relooking at the capabilities of bus and train. The latter must cater to a wider variety of passengers, not just the slim, agile and baggage-less. They must still be comfortable at surge times, not just look good based on statistical averages. And much greater priority should be placed on reducing travel and connection times. If the existing duopolistic companies aren't keen to do this, then we must have new competitors.

The argument may be made that while multiplying routes and increasing the capacity of bus and train may help with the surge problem, they create another problem of excess capacity at other times. Clearly, the extra vehicles have to be mothballed during off-peak periods. Once again, the government can play a part, licensing buses only for peak-hour runs, but in return doing two things: allowing these buses roadworthy certificates many years longer than for normal buses (considering that these limited-hours buses suffer less wear and tear), and waiving various taxes on them. The Manpower Ministry can also permit foreign workers (with lower wages) to be hired as drivers. Perhaps that way, there can still be found an economic formula for tripling or quadrupling transport capacity during peak hours.

Until we do all these non-taxi things, there will never be a solution to the taxi problem. 

© Yawning Bread