Delhi’s Odd-Even Rule Is At Odds With What Needs To Be Done [Part 2]

[I  decided to provide an analysis of Delhi’s Odd-Even Rule in two parts for easier digestion: you can find some background on Delhi’s pollution issue, the rationale of the road rationing scheme and whether it’s working in Part 1 here]

The introduction of Delhi’s Odd-Even Vehicle Rule is unlikely to have the level of impact that many are expecting. But those who had studied the numbers had probably predicted that already.

Let’s briefly put the odd-even restrictions on vehicles into context. There have been quite a number of misleading reports over the last week (some from highly-respected sources) that Delhi will halve the number of vehicles on its roads. That’s unfortunately not true. First of all, the odd-even rule only applies to private four-wheel cars. Two-wheelers, three-wheelers, and trucks are not included (neither are buses but to ban them would be pretty counter-intuitive!). However, even if we said that it would halve the number of private cars, we’d still be wrong. There are a number of additional exemptions to the rule which limit this restriction further. The following are exempt:

  • Commercial vehicles (i.e. taxis-esque services)
  • Women drivers (solo, in a female-only car, or with children under the age of 12)
  • Private vehicles running on CNG (a cleaner type of fuel), as well as electric and hybrid cars
  • VIP escort vehicles (Presidents, Prime Minister, Union Ministers, Chief Ministers, judges, embassy vehicles)
  • Emergency vehicles (thankfully!)
  • People on their way to hospital for medical emergencies will be exempt on a “trust basis.”

In other words, the rule applies to non-VIP males driving a private petrol or diesel car. In which case, it’s slightly misleading to claim it’s “halving Delhi’s vehicles”.

Anyway, if this proportion of drivers comprise a large contribution to Delhi’s air pollution then the restrictions could still be effective. How do the number stack up? I had a hunt around to try to find some data on how Delhi’s pollution breaks down by source (shown below)*. As you can see, transport contributes 22-23% to the city’s PM2.5 concentration. [I should note that the contribution of different sources varies between seasons: in summer, vehicles only contribute roughly 10% to emissions—but since pollution levels are highest in the winter, I’ve used the 22% winter figures]. While we often assume that the main culprit of polluted cities is road traffic, sources such as waste-burning, diesel generators (especially in developing nations), industry, dust and domestic activities are surprisingly large.


Even if the odd-even rule did manage to halve the number of road transport vehicles (and correspondingly reduce their emissions by half), pollution levels would only be reduced by 11% (admittedly this is a decent chunk, but probably difficult to see clearly in amongst noisy measurements). But when we breakdown transport emissions further by vehicle (see below), we see that 4-wheel cars only contribute 4% to total emissions. In this case, even all private cars were included (i.e. no exemptions), we’d only be cutting emissions by a few percent at the most.


Targeting this small slice  is not only going to be deficient in scale, but improvements are likely to be largely outstripped in the long-term. A team of 6000-10,000 volunteers have been drafted in over the trial period to monitor implementation, and driver compliance has been impressive so far. However, such levels of monitoring are probably unsustainable and it’s likely that compliance will fall when this gets more lax. Some drivers could probably find ways of bypassing the rule by registration of different number plates. It’s also important to consider the staggering rate at which vehicles are being added to Delhi’s roads—approximately 1400 per day. From my calculations, if that rate were to stay the same, Delhi would have added around 2.5 million more within five years.

That raises a further concern: will this rate of increase stay the same? True, the restriction might discourage people from purchasing a vehicle, but it could have the opposite effect. For those with money, what’s one of the easiest ways to beat the scheme? Buy a second car with the opposite number plate (and probably an older, dirtier one if it’s your second). Of course, being able to afford one car isn’t economically feasible for many Delhiites, but with a rapidly growing middle class there is the potential to drive many towards buying a second. This has proven to be true in cities such as Beijing, Mexico City and Bogota.

The odd-even rule is a scheme focused on short-term rapid reduction in pollution. Beijing implemented it in the run-up to the 2008 Olympics and at the time it was pretty successful; I still didn’t envy the marathon runners, but they did experience a significant dip in emission levels. Several years down the line and Beijing’s pollution levels are nothing to envy.  However, at the time China also tackled the other key slices of the pie: it halted more than 100 factories and 56 power plants (mainly coal-fired) for the duration of the games. It also invested billions in its public transport (especially its metro and bus systems) to cope with the displacement in private travel.

This is exactly why Delhi needs to do the same, but with a long-term infrastructure focus. In terms of transport, the city needs a major focus on the improvement of its public transportation systems; it needs to make it favourable and attractive for people to choose public over private transport. That’s not going to happen if these options remain unreliable, overcrowded and unsafe. That’s not going to be an easy or cheap task, but it’s one the city needs to face, especially in light of its rapid economic and urbanisation growth rates. It’s also clear that focusing on transport won’t be enough to clean up Delhi’s air. Focusing on a single 22% slice is insufficient with pollution levels 16-20 times the recommended safe limit.

Despite my criticism of the trial, there are some positives to take from it. It has, if nothing else, drawn significant attention both nationally and internationally to the scale of the city’s pollution issues. There are clear signals that this is an issue that desperately needs to be addressed; the overwhelming response from Delhi citizens is another signal that the public pressure for change is strong. The reduction in traffic congestion has been another significant positive—many have noted the increase in productivity that reduced travel time could help.

It’s a start nonetheless, but falls well short of what’s needed. We don’t need to look at air quality measurements to see this—we could have predicted it from some background numbers. Ultimately, if you only tackle a little, you’re only going to achieve a little.


*Although there are small variations in quoted numbers from different sources, I’ve taken the most recent (2014) study data from the Central Pollution Control Board; CPCB (which has been widely cited) which I’d expect to be most legitimate. I haven’t seen any figures which vary hugely from these, so I’m going to assume they’re pretty close (or close enough to use for back-of-the-envelope context).


Delhi’s Odd-Even Rule Is At Odds With What’s Needed [Part 1]

[I have decided to provide an analysis of Delhi’s Odd-Even Rule in two parts for easier digestion: the first to provide some background on Delhi’s pollution issue, the rationale of the road rationing scheme and whether it’s working; the second to provide some further analysis, statistics and reflections on why it probably won’t have the level of effect Delhi desperately needs]

You’d be forgiven for thinking that Beijing was the world’s most polluted city—on the subject of air pollution, China’s capital consistently steals the headlines—but according to the World Health Organisation (WHO), Delhi actually claims the top spot. There are a number of important air pollutants (and Delhi scores badly on all of them), but that of most concern is PM2.5 (particulate matter measuring only 2.5 micrometres in diameter. PM2.5 comprises tiny particles of dust, soot, dirt, smoke, liquid droplets in the air. The size of these particles means they can easily penetrate the lungs, increasing the risk of a range of cardiovascular health issues including cardiovascular disease, lung cancer, bronchitis, and heart attacks. In many cities air pollution is a fatal issue, estimated to cut years off the lives of its residents.

Comparison of Air Quality in Delhi and Beijing 2014-2015

Delhi’s concentrations of PM2.5, like most growing cities, have been on the rise. The figures are pretty shocking. The city’s concentrations have recently been consistently reaching 400-500mg/m3—a staggering 16-20 times the WHO’s “safe” limit (25mg/m3). To try to combat Delhi’s air pollution issues, officials launched the trial of an odd-even road rationing scheme last week (starting January 1st and running for two weeks). This scheme, which has been tried in other cities, aims to cut road traffic by regulating private cars to travel every second day. Essentially, it’s implemented by your car’s number plate: if your plate ends in an even number, you can only drive on an even date, and vice versa for odd number plates. The rationale works on the basis that if you take a large proportion of emitting cars off the road, you’ll remove a large proportion of city’s particulates.


Has the scheme been working? The public response has been impressively cooperative (maybe not so surprising when you consider how debilitating the issue must be for Delhiites)—the number of drivers caught violating the restrictions has been in the 100s-1000s, which is small for a city with roughly 9 million vehicles. There have been a number of positive reports from locals on congestion and traffic improvements on city roads. But has it had an impact on air pollution levels? The government plans to publish full results at the end of the trial period, but from the reported numbers from various sources so far, the opinion was very much split. Some have reported that the scheme is working, others report that pollution levels are actually climbing.

The truth is that none of the comparisons are particularly valid. Air pollution levels are incredibly variable [see image above] and dependent on factors such as the season (winter is typically worse than summer), the meteorological/weather conditions (wind speed, direction, humidity, rainfall are all important), the time of day, the type of measure you’re using. One source reported a dramatic improvement…by comparing December’s peak reading to today’s average reading. That doesn’t tell us much more than the fact that peaks tend to be larger than, well, peaks. It’s data cherry-picking for a favourable headline. Other sources reported an increase since the restrictions were put in place, but seemed to forget that they were measuring during a period of very low wind and high humidity (particulates are basically clinging to water droplets in the air with no wind to disperse them). Their comparisons are no more valid.

In other words: air pollution is a complex issue that you can’t properly draw conclusions from using only a couple days’ data. To realistically evaluate the impact of the scheme, we’d need to compare like-for-like over a much longer-term dataset to average out the variability. What we can conclude, however, is that it’s probably not going to provide the scale of improvement Delhi desperately needs. To be debating whether air pollution levels have gone up or down a little when Delhi’s pollution is 16-20 times the recommended safe limit pretty much illustrates this point. However, we don’t need to wait for long-term direct pollution data to understand this—we can work out why that might be the case by putting some more general statistics and considerations into perspective.