Making odd-even work

The announcement by the Delhi government that, as a measure for improving air quality in the capital, odd- and even-numbered private vehicles would be allowed to ply on alternate days of the week from January 1, 2016, has given rise to both hope and apprehension.

Hope that reduction in the numbers of vehicles on the road would lead to reduced emissions and traffic congestion. That the move will strongly incentivise car-pooling and other demand aggregation arrangements, leading to a long-term solution to Delhi’s transportation and pollution challenges.

Apprehension that, first, some citizens could game the system by acquiring old polluting vehicles with number plates complementary to that of their existing vehicles, or acquiring quick-change false number plates, so that there is no actual reduction of vehicles on the road, and possibly generation of even more pollution. Second, that some citizens may collude with traffic policemen to violate the odd-even rule. Third, that public transportation services cannot be augmented sufficiently or rapidly enough, so that in the absence of adequate alternatives, the scheme would collapse. Each of the grounds for hope, as well as of apprehension, are clearly valid. How does one ensure that the hope is realised, and that the apprehensions are contained? There are some suggestions.

First, false number plates. The traffic police could adopt a randomised inspection regime (identifying vehicles for inspection on the road on the basis of computer-generated random numbers) to match number plates with registration certificates. If false, the owner should be charged with the full penalties applicable under the Motor Vehicles (MV) Act. How does one prevent collusion between black sheep among the citizens and the police? By designating a large number of spots for roadside inspections, equipped with real-time CCTV cameras, monitored perhaps by the anti-corruption bureau. The traffic police teams would need to be stationed randomly (unpredictably) at different inspection sites to reduce chances of prior collusion. Delhi Police has also put out a mobile app to enable citizens to report traffic violations, which may further help improve enforcement of the odd-even rule.

Second, acquisition of old vehicles with complementary number plates. By creating an environmentally benign demand for old vehicles, i.e. requiring that any purchase of a new vehicle would require the scrapping of an old vehicle of greater emissions burden. This would also stabilise the total number of vehicles in the capital, and improve emissions levels on average. Would this be unfair to first-time vehicle owners? Not significantly, if electric vehicles are exempt from this requirement.

Third, that publicly provided transportation services cannot be augmented rapidly, given accountability requirements for the public sector (CAG, CBI, CVC), or sufficiently, given budgetary limitations. This would require policy-makers to discard one hoary shibboleth-that transportation for the public must be provided solely by the government. However, involving private service providers, from single vehicle-owners to large corporate fleets, requires one to understand the nature of urban transportation demand, and to design a competitive market for such services.

Travel demand by urban dwellers varies across the time of day, day of week, month of year, and also changes over the years. Most of this demand is patterned, i.e. there are definite regularities as regards origin-destination pairs, time of travel, comfort levels, acceptability of fellow travellers, and willingness to pay. Some travel is unpatterned, such as a sudden emergency visit to the hospital. Unpatterned demand can only be met by personal vehicles or taxis/autos. However, patterned demand, even where the patterns are complex, can be met by private providers, if they have the necessary monetary incentive to collect, and act upon, information about such patterns. Such information is difficult for government-owned transporters to collect and act upon, given institutional rigidities. Information technology can immensely help private providers to gather the necessary information, even in real time, and some aggregators have started providing services.

How then does one operationalise a competitive, privately provided market for urban public transportation, distinct from individualised taxi services?

First, private service providers at different scales, from autos to buses, and including cars, vans and mini-buses, may be registered, without limit on their numbers, to provide urban transportation services. For this, they should have two certifications. One, the vehicles must be safe, insured adequately for passenger and third-party injury, and meet emissions norms. These vehicles should have prominent identification. Two, the service personnel (drivers, conductors, etc) must undergo police verification. The drivers should also have valid, clean driving licences. They should be issued badges to be displayed on their persons that they are certified.

Second, a very large number of pick-up and drop-off points may be identified for transportation providers. Ideally, none should involve a walk of more than 50 metres, i.e. they may be no more than 100 metres apart. Beyond this distance, commuters may be disincentivised, especially given Delhi's severe climate. On main roads, for each class of registered vehicle, a minimum occupancy should be specified—for example, 4 passengers for cars, 12 passengers for mini-buses, etc. This would help ensure actual diversion of passengers from private cars to registered service providers.

Third, no further restrictions may be placed on origin-destination pairs (as long as they are designated pick-up and drop-off points), timings, comfort levels, service pricing, or profile of clientele (i.e. if service providers wish to restrict their services to persons of a certain profile, there should be no objection to their doing so).

These steps will, given time, help build a competitive, differentiated market for urban transport. A control-freak mindset of policy-makers seeking to place further restrictions will kill it.

All of this will require preparation, may involve some legal amendments, and perhaps January 1 is too early for a start. An ill-prepared venture would result in a public outcry, leading to failure. Also, a standalone solution for Delhi may be problematic—a coordinated policy across the NCR would be advisable.