Tuesday, May 11, 2010

Kerala's highway widening controversy

While in Trivandrum for a few days, I came across the raging controversy about what should be the width of the National Highways (NH), and could not resist wading into the debate. These quick comments about the specific situation in Kerala is from a cursory understanding, though its general thrust is relevant for all road-widenings.

The crux of the issue is that while the National Highway Authority of India (NHAI) has proposed widening NH 17 and 47 to a width of 45 m, public opinion (crystallized in an all-party meeting) favors restricting it to 30 m.

In this context, I am reminded about an old post, in the context of Vijayawada road-widenings, that argues in favor of framing road widening debates in terms of lanes rather than road widths. I am inclined towards the view that the debate should have been formulated as a requirement of 3/6 (or 4/8) lane carriageway with a 4-5 m wide service road, instead of the 30-45 m width choice.

Taking cue from behavioural psychology, it is far easier to bargain for a reduction in road width by 5 or even 15 m, as opposed to negotiating for reduction in one lane or dispensing with the service road. One carriage lane or a service road has a much greater cognitive salience than 5 or 15 m. Anybody who has done road widenings on scale would appreciate this dimension of collective psychology.

In many respects, road-widening is a classic collective bargaining game where framing the context of the negotiations is critical. An appropriate formulation of the choices can considerably increase the odds of achieving a successful result. Therefore, and in view of the aforementioned political dynamics, a more realistic final goal should have been to push through atleast a six lane carriageway proposal, even if without any service lanes.

Given the fact that existing roads are 10-15 m wide, a 45 m road widening proposal would have meant an effective tripling of the existing width, a difficult proposition in a densely populated and urbanized context, as in Kerala. At best of times Kerala is a difficult place to push through such measures. The numerous examples of tortuously slow road-widenings in Trivandrum city is a reflection of this difficulty.

In a state where travelling 60 km on NH 47 from Trivandrum to Kollam takes more than 90 minutes, the impact of even a 30 m road widening on Kerala's economy can be dramatic. To the extent that road transportation is one of the critical infrastructural engines of economic growth, substantial economic activity in the state remains suppressed/disincentivized by the narrow highway roads.

A 30 m (or 100 ft) road will have 1.2 m (5 ft) central median, with 10.5 m (=3.5x3) (or 33 ft) of three-lane carriageway, about 2.5-3 m of drainage, and 0.6-1.1 m of shoulders on either side. This specification would be adequate to meet the basic requirements. All the more so, since unlike other states, Kerala does not serve as a transit pathway for the North-South or East-West passenger and freight traffic. Further, given its limited industrial base, a wider NH would be more useful in facilitating the growth of the state's services sector than promoting freight traffic. While a 60 m width would have been a luxury, and 45 m width ideal, the aforementioned practical considerations means that a 30 m width should be a good second-best solution.

Further, the success or otherwise of road widenings are critically dependent on the flexibility in the widths across stretches (without compromising on the essentials) and compensation payment strategies. This is all the more so since given the urbanized nature of the state (especially adjoining the NH), the NH widening across a major length (and from where resistance is likely to be most vocal) will be similar to widening of city roads. In such environments, decentralized negotiations (at official level, as is done in states like Andhra Pradesh) can yield quick and fairer results, though it carries the risk of politicization and corruption.

Though, in comparison to the regular land accquisition process, the NHAI's procedures are faster and its compensation amounts larger, such urban land accquistion will have to embrace more innovative approaches like transferable FAR bonds (TDRs), waival/concessions on building fees, impact fee concessions etc. Further, there should be a distinction made between those losing the major share of their property (as to leave the remaining extent virtually useless) and those who lose only a smaller share of their property; houses and commercial establishments; and owners and tenants.

There is an Econ 101 arguement in favor of the 30 m proposal. Even with the 45/60 m widening, the carriageway will be no more than 4 or 6 lanes, and the remaining land will be left barren as shoulders to be brought to use when required. In any case, the NHAI roads do not have any drains and the concept of ducting to lay utility lines is yet to catch up even in the metropolitan cities.

In the circumstances, we are looking at accquiring and leaving 15 m (or 50 ft) of road width for a length of 500 km (i am not sure of this figure) virtually unused for a considerable period of time (let us say 10 years). Leave aside the virtual certainty of its encroachment, especially in the densified and urbanized context of Kerala, the sheer economic waste incurred by leaving 7.5 million sqm (or around 16000 acres) of valuable real estate (in fact, the most valuable lands in the respective towns/villages) area unproductive is staggering.

Back of the envelope calculations assuming a very conservative rate of Rs 5 m per acre would yield an asset worth Rs 80 bn (or Rs 8000 Cr) proposed to be left idle. The net present value of this investment, over a period of 10 years and assuming a discount rate of 10%, would surely be many times more than the cost of accquiring the same land after ten years. This calculation ignores the net economic return components like tax revenues from any prospective economic activity (which would be imminent once the widening is completed) on these lands and the contribution of the local economic multiplier due to this activity.

Even assuming sky-rocketing land values with road widening, there is surely some broad limit beyond which land values cannot just shoot. Though these calculations are also valid for other states, the virtually rural and uninhabited nature of their (say Andhra Pradesh or Orissa) NH stretches means that the opportunity cost of leaving them idle is not very large.

If the 30 m width is finally agreed to, the government should immediately demand consent letters from all those affected, and the demolitions should be done at the earliest. The momentum generated by the width concession should be leveraged to complete the widening and construction in quick time.

In light of this, a more practical choice facing the government now is between a swiftly executed 30 m widening and a lingering 45 m widening entangled in endless litigation. Take your pick.


Ajaypp said...

I agree with most of your points, especially the kind of hurdles faced by project land acquisition, the approach of presenting the widening in terms of lanes instead of mere width and also the fact that it is essential to distinguish between people losing the major part of their property and those losing just a minor part (they may benefit due to the appreciation of property post widening).

However, there are a few points to be noted:

1. The whole premise of the 30 m argument is based on the fact that we are looking only at 4-laning. However, the plan to 4-lane was based on traffic studies done in 2003-4. Today, NHAI itself accepts that the traffic on the NH-47 has exceeded the threshold for 6-laning.

2. Even with a 4-lane road, once we consider essential components like the service lanes and utility corridors (necessary to prevent the highway getting frequently dug up to lay pipelines and cables, the 30 m RoW proves insufficient.

Let's take 4 X 3.5 = 14 m for the 4 lanes, 1.5 m for the median, 2 X 1.5 m = 3 m for the paved shoulders, 2 X 1.5 = 3 m for the unpaved shoulders, 2 X 2 = 4 m for the utility corridor and 2 X 7 = 14 m for the service lanes. This works out to 1.5 + 14 + 3 + 3 + 4 + 14 = 40 m. This leaves scarcely 5 m for future development, meaning that an additional 2 lanes can barely be squeezed in, perhaps by trimming the shoulder width.

3. Service roads are essential in Kerala as opposed to the intercity highways in the US, UK or other Indian states since the roads pass through urbanized and semi-urbanized areas where there are lots of side-roads and heavy local traffic. Without service roads, the Highway would become congested and accident prone.

4. The NH-47 in particular has become the arterial route for both passenger and cargo flow in South Kerala. This will become even more pronounced as Trivandrum rapidly expands and massive projects like the Vizhinjam deep-sea port, the Trivandrum IT Corridor and the Cochin port develop along its length.

5. The State and local Governments have to conduct a detailed social impact study to identify the real effects of land acquisition and the most seriously affected people.It's also their responsibility to work out a fair compensation & rehabilitation policy and ensure its implementation along with NHAI and GoI. This cannot be shirked by curtailing the acquisition.

Look forward to hear your thoughts on this.


gulzar said...

thanks ajay

1. i am not sure about 30 m being a four lane road. a 30 m road, with 1.2 m median, can easily be a 6 lane (21 m) road - there is enough left to accomodate small shoulder and drains. all 100 ft city roads are six lane ones (the covered drains serving as footpaths). but there cannot be any service road with a 30 m alignment.

2. as i had indicated in my post, we are trying to make the best out of a really challenging situation. so, instead of the ideal solution, we ought to be trying to adopt a second-best solution. and it becomes all the more acceptable, when the NHAI itself is not doing any duct (except for a few miniscule stretches, none of the major indian cities have ducts, and ducts have their problems... but that is another issue) across its stretches and when the unpaved shoulders can be used to lay pipes/cables.

as i mentioned earlier, service roads become a choice we have to give up with 30 m. Or else should try to have specific stretches, around really dense local traffic areas, with service roads. there is surely no need to have service roads all along the length of the NH.

3. i am a little sceptical about this service road reducing arguement. as far as possible, NH should bypass dense urban centers. if NH passes through such areas, no matter what types of access controls and service roads are in place, the local traffic will inevitably spill over. we surely cannot have flyovers and underpasses criss-crossing these roads to channel traffic. even in areas with service roads and access controls (just travel any stretch, adjacent to small junctions/towns abutting the NH), traffic spill-over into the main carriageway is commonplace.

Ajaypp said...

Thanks for your thoughts on these points. Let me go point-by-point:

1. Trying to fit 6 or 8-lanes within the so-called 30 m RoW is NOT the responsible approach for a Government. After all, why is the figure 30 m and not 25 m or 35 m. There seems to be little scientific basis for it and it could get trimmed even further. A Highway without the basic components won't be a Highway at all and will not serve the purpose that it is being created for at such massive expense.

2. Yes, we are looking for a pragmatic solution but I believe one was found at the earlier agreement to have a 45 m RoW. What is to stop another "all party" consensus from calling for the RoW to be reduced to 20 m just before the next round of elections?

Utility corridors are more relevant in Kerala than in other States because a lot of utilities are routed along the NH due to a lack of RoW elsewhere. For example, the State's optic fiber backbone lies along NH-47 and I am sure future utility expansions like the proposed natural gas pipeline system may also opt to follow suit.

As for the shoulders, I agree that they can be curtailed or dispensed with in the less urbanized stretches, but the debate is mostly about the more urbanized stretches where the service lanes are needed.

3. The NH-47 is already bypassing all three major urban areas along its path - Alleppey, Kollam and Attingal. However it cannot bypass every minor urban area without major realignment which would take decades. Also, many of the commercial establishments in these small towns would depend on the NH for much of their business.

I agree that service lanes would not completely prevent spill-over of local traffic into the main carriageway, but it will reduce it to a great extent.