Syndicate content

Life in the Slow Lane - The Nairobi Grind

Apurva Sanghi's picture

I’ve lived in cities famed for their gridlock: 1990s Bangkok (gridlock was as bad as it could be); Los Angeles (gridlock + pollution); New Delhi (gridlock + pollution + honking galore); Nairobi’s gridlock is surely up there.

But is traffic “bad”? What sort of question is that you ask? Surely, the answer is 'yes', you say: time wasted stuck in traffic, the frustration, the needless idling of vehicles which creates both local (and global) pollution and so on. But let me suggest this: traffic congestion is also a sign of development. In fact, the more vibrant and dynamic the city as Nairobi surely is, the more traffic congestion you might paraphrase Gordon Gekko from the movie Wall Street, “Traffic is…good”!

Let’s face it though – you and I don't like to be stuck in traffic – indeed, who does? So how can one crack the traffic congestion problem in a city such as in Nairobi? Let me touch upon three common -- and one uncommon -- solutions:

First, the infrastructure solution: Why not build more roads, highways, bridges etc.? Isn’t that an obvious one? Perhaps, but the problem with this approach is the “more you build, the more they’ll come”.  Traffic congestion eases temporarily and it is only a matter of time before more vehicles are added to the road to offset this initial easing (over 10,000 new cars are apparently added to Nairobi roads each month)! Now this is not to diminish the importance of adding new infrastructure when needed: for example, if somebody from the north of Nairobi wants to travel to the south, building bypasses that do not require commuting through the central business district makes complete sense. Indeed, Nairobi has just built about 100kms of new highway (some with bicycle lanes). But as mentioned above, such solutions tend to be “bandaid” ones (i.e. temporary). And the more economically vibrant a city like Nairobi, the less effective this solution.

Second, the public transport solution: Why not put more public buses (or allow private ones) on the road? Why not build a metro rail transit system? While this seems commonsensical, my original home city, Delhi, added a metro system only as recently as a few years back. My last home city, Washington DC, has yet to complete a public metro train line that links the nation’s capital's international airport to the city! And my current home city, Nairobi, does not have a metro rail system. It does, however, have customized, private mini buses, called “matatus” (which as an aside, are an artistic and cultural phenomenon of their own). But these too are victims -- and contributors -- to gridlock. There is a growing market for boda-bodas (inexpensive motorcycle taxis) but they are not yet as reliable as the motorcycle taxies (“motos”) I relied on living in Bangkok: zippy and inexpensive two-wheelers that weave and whiz past stalled luxury BMWs and Lexus-es, guaranteed to get to your destination pretty much on time (just hold on to your knees though)! So why don’t we have more public transport? Metro rails? Motorcycle taxis? I do not know but one typical reason elsewhere is the local political economy -- existing vested interests often block the emergence of new, efficient forms of transport. And this is not just a developing country phenomenon – when I moved to Los Angeles in the late 1980s for study, there was no public transport whatsoever (and I had to buy a car out of necessity). You see, historically, the Big 3 American automotive companies systematically blocked the emergence of public transit in California because of their desire to sell more cars!

Third, the information solution: Ever-innovative and enterprising Nairobians have begun marrying traffic and technology, and have even developed an app for it. “Ma3Route” is a mobile phone app that crowdsources traffic information. Linked to twitter, it provides live feeds to its subscribers to help them avoid / bypass local bottlenecks. In addition, an IBM team has put forth a plan of using readily available technologies such as cell phones; CCTV; GPSs in Nairobi etc. to make traffic information more readily available both to drivers and traffic police. That can help pinpoint traffic congestion bottlenecks in real time and better manage congestion, and the Nairobi county government is considering adopting key proposals of this plan. A common feature of this approach is providing live, accurate information to permit real-time trip scheduling / routing. My feeling is that these improvements in information help but mostly on the margin…so the question remains: how to make a big, systematic dent? Read on...

To make a dent, here's the “uncommon” economists’ solution that some of you might be familiar with. It builds on the economist’s zeal of improving complex social systems / behavior by changing incentives, in this case, through the price mechanism. Why not simply charge people for the costs of congestion they impose on the system? Urban transport experts tell us that traffic congestion is “non-linear”, i.e, once traffic goes beyond a certain threshold even by a small amount, the cost of congestion (say, measured in delays) shoots up exponentially (intuitively, because of driver reaction time delay. For a slightly more technical explanation, click here). This non-linearity then suggests that discouraging even a small number of people to not drive during peak rush hour can lead to big improvements in traffic congestion. How do you do that? Start charging a small fee (called “congestion pricing”) in peak rush hour. Now this was first experimented in one of my adopted home cities, Singapore. In the late 1990s, Singapore introduced the Electronic Road Pricing (“ERP”), which is a usage-based, real time, road pricing system. And it did precisely that – it starting charging drivers a small fee based on the congestion they impose on the system (so higher fee during rush hour and vice-versa), and in conjunction with the three common solutions above (decent infrastructure; reliable public transport; making use of available technologies), pretty much solved its gridlock problem. Having lived in Singapore and experienced it firsthand, it is indeed a real marvel! And this has taken off in other places. Toronto, and more recently, London and Stockholm have been inspired by the Singapore model, and have introduced congestion pricing to much success. Now you might think that people would object strongly to pay for something that used to be free. And you'd be right -- when first introduced, Singapore’s ERP was nicknamed “Everyday Rob People”. Not anymore though. Similarly, what happened in Stockholm is illuminating. There was the initial, expected resistance but now a majority of users so strongly support congestion pricing that it being extended to Sweden’s largest city -- Gothenburg. Of course, these solutions are not mutually exclusive – you do need to build and maintain roads and you do need decent public transportation alternatives. But good old-fashioned incentives work on us humans and can do magic. How to make them work for, um, nonhuman species that cause gridlock in Nairobi is a completely different story…please see below… 


Submitted by CAA on

As a Nairobian living abroad (in a city with excellent public transport, bicycle lanes and small enough so that you can walk anywhere at anytime, day or night), I have often thought that charging people for congestion is an excellent idea. That is, until I visit Nairobi and I remember that in certain parts of the city, walking after dark is a problem. Taxis are not regulated, and so depending on weather, time of day or night and so on, it is probably cheaper to drive if you have an engagement that ends after daylight. Perhaps solving the insecurity problem first would certainly act as an incentive to take public transport or cycle or walk, in addition to introducing charges for congestion.

Dear CAA: Thanks for your comment, and I agree with you that addressing the insecurity issues on Nairobi roads would incentivize the use of the public transport. In fact, my observation is that insecurity (which also stems from poor / absent street lighting) makes people -- for those who can afford it -- buy bigger and more expensive cars, which then impose external costs on others through increasing congestion and worsening the quality of existing roads! So a low cost intervention such as improving street lighting could go a long way in addressing many problems...
PS: Even after a year living in Nairobi, I am surprised that taxis are not metered, and there is no action towards that either. That is not surprising because taxi drivers are typically a strong vested lobby. Not having meters may well have a progressive outcome, in that, taxi drivers are incredibly astute in discriminating clients based on their ability to pay -- so richer passengers get charged more and fact, I've seen quite a few instances when taxi drivers give short rides for free to local / domestic workers.

More information about text formats

•Allowed HTML tags:

•Lines and paragraphs break automatically.

Notify me when new comments are posted

All comments

Replies to my comment

By submitting this form, you accept the Mollom privacy policy.

Submitted by Sosthenes Bichang'a on

I am the President of the IICFIP Inc ( who stays in Nairobi. Traffic here is a nightmare especially at mornings and evenings but I appreciate the joint effort of County and National Government in expanding the roads network to ease congestion.

Submitted by Apurva on

Thank you for the comment Sosthenes. You are right that there are efforts being made to ease congestion but as I note in my blog, while expanding the road network is essential, that approach falls victim to the "more you build, the more they'll come". Especially given the rate at which new cars are being added on the road. Hence the suggestion for congestion pricing (in addition to expanding the road network). Best, apurva

Submitted by Mueni on

The rapidly growing car population is a big challenge that may need to be dealt with by making it expensive to import the second hand cars. I like the Everyday Rob People solution but it would need to go hand in hand with labour policies that would allow for staggered work hours. In addition,implementation of the 24 Hour Economy strategy might provide a solution to the Nairobi grind.

Dear Mueni: Interesting suggestion for staggered work hours. De facto, I think that is already happening in many corporates / private sector, with employees beign given more flexibility around working hours because of the traffic situation. Same goes for being a "24 hour" economy, as many of us, in an increasingly globalized world, given clients in different time and spatial zones, work around the clock anyway. (Not so sure how the 24 hour economy would apply to agriculture though, which still comprises a whopping quarter of Kenya's economy) Best, Apurva

Submitted by Mueni Kimulu on

Dear Apurva.. I was talking of 24 hour economy in the urban setting specifically to deal with the gridlocks experienced in Nairobi through different shifts which would flatten the traffic.

Submitted by Apurva on

Thanks for the clarification Mueni. Given how services based the nairobi economy is, I think that shift is already happening if not by design, then default! Cheers, Apurva

Submitted by Shanta on

Apurva, thanks for this delightful piece. You might add that congestion pricing is correcting an externality--the cost that each additional driver imposes on other drivers. So it is not that people now have to pay for something that was free, but rather that they have been subsidized all this time, and now they have to pay their fair share. The absence of a congestion charge is as much an unproductive (and possibly regressive) subsidy as are fuel subsidies. So the people advocating for reform of fuel subsidies may want to join forces with those advocating for congestion pricing. In both cases, we should protect poor people with cash transfers.

Thanks Shanta. Agree, and I couldn't explain the point any better than you did! And great point linking the issue to fuel subsidies and cash transfers. Cheers, Apurva

Submitted by John Wille on

Frankly I think the situation is hopeless unless they can reform the traffic police. Nairobi installed traffic lights all along the Uhuru highway and the cops ignore them and still control the flow of cars... Go figure...

Thanks John. Sure, reforming traffic police is part of the fitting the puzzle together but until then, why not try congestion pricing that is less prone to petty corruption? Best, Apurva

Submitted by Daniel Gerber on

I am wondering, if indeed we are worried about the well being of the bottom 40% in a free market economy I am not so sure that this commuter tax model really is the panacea presented in this article. The most expensive real estate per square meter is downtown where the high income people live, while the less moneyed tend to live in the periphery. The question then is why do we tax the poorer who commute from the periphery to come to work rather than the wealthy that live downtown and do not really need to commute. I understand that Japan has a tax on vehicles depending on where you live, with downtown vehicles being subjected to much higher tax rates than vehicles outside of city centers, this concept could be expanded to property taxes, even VAT. Of course, this is not such an easy question as there are costs to decentralising service provision of infrastructure too. But just charging a commuter tax to those that come from the periphery just favors those that already benefit from living downtown.

Dear Daniel: Thanks for your comment. If anything, I would argue the opposite. Who typically drives cars? The relatively well off. But they impose an external cost to the poor / pedestrians (through congestion; poor use of public transport; pollution etc). So by taxing the relatively well off through congestion pricing, you're actually improving prospects for the relatively worse off! You can also think of a differential pricing scheme where public transport (used by the relatively well worse off) is exempt from congestion pricing, thereby increasing the incentive for using public transport. Cheers, Apurva

Submitted by Daniel Gerber on

Dear Apurva - I made the mistake of not naming cities I was referring to. Indeed in most developing countries only the wealthy have cars. I was more thinking of the discussions taking place in Western Europe or even here the US where the majority of people do actually have a car and only the very wealthy can afford to live in the heart of the city. Middle income and poorer people in these places live more in the periphery. Maybe the problems are not quite the same everywhere.

Submitted by Apurva on

Indeed Daniel. Agree that problems are not the same everywhere. Although I'm guessing that what you say about middle and high income countries is true for Stockholm and London but congestion pricing seems to work v well there. Cheers, Apurva

Submitted by Anonymous on

The wealthy in Nairobi do not live downtown...Nairobi has minimal downtown residences. However the point about the bottom 40% is important....whatever measures are used they shouldn't make the poor more poor. Doubling the cost of parking has not made a difference so it should target stalling car population growth. Improving public commuter services might be attractive but this will require major reforms and infrastructure!!

Submitted by Apurva on

Dear anon: agree with your objectives but perhaps differ on the means. As people grow richer they will want more cars. So doubt if stalling car popln growth will help. And regarding the bottom 40% I think my response to the previous poster still stands. Cheers, Apurva

Submitted by Kathy Whimp on

Back in the 1990s on my first visit to HQ (I was an IFC STC then) I stumbled across a fascinating book on the completely unregulated practice of informal/casual carpooling, called slugging. Here's a website about it:
The incentive that drives this access to the high capacity lanes that can only be used by cars with three occupants.
What interests me is the administrative systems needed to "enforce" the incentive. After all, a high capacity lane only works if there is a genuine possibility that drivers will be charged for using one incorrectly. In Australia this works through the vehicle registration system, which permits virtually automatic penalisation of anyone detected (by an automatic camera system that records the vehicle registration)in a fast lane without the right number of people in the car I assume that's how the ERP system in Singapore works too - charges can be automatically billed to a vehicle's registered owner if they drive into the area without the tag that allows the ERP charge to be automatically charged to their account? And what does it take to make such a system leakproof? In Australia I know I have had huge trouble getting my drivers licence renewed since I moved to Kenya, because only residential addresses are permitted to be associated with drivers licences, and proof of residence in the form of a lease or utility bill is required to register a new address. That's because the system needs to be able to follow up road-related penalties (like driving through a red light or a speed camera, or failing to pay a toll) with service of a court summons in the event that I fail to pay. In Nairobi we don't have street addressing yet - post only goes to post office boxes. Are there countries where these systems work without the rolls royce underpinning of linking individuals to locations for the purposes of enforcement
By the way, on a recent visit to Jakarta I was introduced to the developing country version of slugging: people spend their days riding back and forth for a small fee, so the driver can access the high capacity lane...

Submitted by Apurva on

Hi Kathy. Thanks for the info. On your query, any car is legally registered to *somebody* even in nairobi. Ghost workers, yes, ghost cars, no :) So I don't think that's a constraint. On slugging, it is very much a developed country thing too -- in fact many of our world bank and IMF colleagues, and others who work in downtown DC, use "slugs" to commute to DC from Northern Virgina! Cheers, Apurva

Submitted by Kathy Whimp on

Perhaps I didn't explain myself very well...

The incentive of a congestion charge works because most drivers believe that if they don't pay the relatively small charge at the time the vehicle enters the regulated area, the penalty for not paying will be much, much larger and a great deal more hassle. To make drivers believe the larger penalty is a real threat, the system needs to be able to ensure that non-payers are pursued, in probably 90-95% of cases, so the word gets around.

My point was not that vehicles are not registered to an actual person, but rather that the registration system records the owner's address as a postal address, rather than the address where they physically live.

So here's why I think linking a congestion charge to a vehicle registration system with PO Box addresses won't provide an effective incentive for very long: Lets say a vehicle goes into a high congestion area without an electronic tag. The automatic cameras are triggered, the vehicle's registration plate is photographed, and the system generates a penalty notice which is sent out to the registered owners address. All that happens relatively automatically. What then provides the incentive for the owner to pay in response to the notice received through the post? The driver has to believe that he or she will be chased down and prosecuted if they don't pay. The usual next step if a driver failes to pay the penalty is to issue a court summons. That has to be served on the person personally. If the vehicle is registered to the owners physical street address, that is relatively easy. If they disregard the court summons a warrant will be issued for their arrest, and the police or a process server knocks on their door to execute that warrant about 7 pm as they are stitting down to dinner.

So what happens in Nairobi (where all the authorities know is the owner's post box number) if the naughty car owner doesn't pay in response to the notice they get in their post box? Following up to serve a court summons on the recalcitrant driver(or arrest them if they fail to appear) becomes a whole universe more difficult and expensive and the congestion management authority is likely to give up at that point. Before long the word gets around that there are unlikely to be consequences from entering a reguated zone without paying the congestion charge.

What makes the congestion charge an effective incentive is the certainty of being prosecuted for non-payment. If the system cannot guarantee to prosecute in 95% of cases, drivers (like 40% of those who receive a property tax bill in Nairobi) will soon learn that not much is likely to happen if they disobey, and the incentive loses much of its power. My point is that these systems of enforcement depend on vehicle registrations including the owner's physical street address (rather than just a PO Box).

Now, I look forward to you telling me why I have got that ALL wrong...:)

Submitted by Apurva on

Dear Kathy. Yes, agree that there needs to be some mechanism for non payment. Perhaps outsource that responsibility to the private sector! Best, Apurva

Submitted by Neville Otuki on

The piece actually represents the thoughts of many Nairobi residents who have to endlessly endure traffic snarl-ups. This is despite there being neat designs on paper by policymakers meant to inject order and sanity back to the urban commute system. Grand plans but less willpower for execution.

Add new comment