Vivek Moorthy: Using roads efficiently- isn't it high time?

With a stiff Vehicle Area Levy, the amount of road area that needs to be built for more travel will fall sharply.

A major failure of our economic liberalisation strategy... is the unrestricted use, coupled with permission for increased production of private cars

""Meanderings of a Gourmet Economist, Review of The Price of Onions, by Ashok Desai, The Deccan Herald, Bangalore, August 6, 2000.

The word infrastructure is probably the most common in economic discussions these days. It is universally stated that our needs for infrastructure are staggering. The new buzzword is "infrastructure deficit". The World Bank has recently coined the term 'infrastructure camel' to denote sectors that export with very little infrastructure"" such as software""and manage to flourish in India's "infrastructure desert." One wonders which sectors in India qualify for the status of "infrastructure hippopotamuses"!

Infrastructure is certainly the biggest proximate hurdle to growth today. (To unearth the deeper factors underlying this constraint would require going into matters of law and economics, not possible here.) However, in some respects, talk of the "infrastructure deficit" has the wrong focus. It is necessary to divide infrastructure into its different components "" ports, electricity, containers, road, rail, air and so on. While there is a huge need for most types of infrastructure, nevertheless, with regard to road infrastructure, there seems to be a staggering deficit of thinking about the subject. Most discussion in India is about building or financing road infrastructure, but unfortunately, next to nothing about using it efficiently.

Inclusive Google searches ("within quotes") on both the Web and from India threw up a number of entries. The citations are broadly illustrative, since by choosing slightly different strings of relevant keywords one may have ended up with quite different results. Looking at the precise string of keywords chosen, the table indicates that in India there has been hardly any discussion about using roads efficiently. 
 

INCLUSIVE KEYWORD STRINGS "" NUMBER OF GOOGLE CITATIONS (AS OF JUNE 15, 2007)
	Whole Web	From India
Financing Infrastructure	129,000	1,790
Efficient Use of Infrastructure	28,000	2
Financing Roads	806	32
Efficient Use of Roads	19,000	1
Road Congestion Charging	9,410	1

Ironically, there are vastly greater constraints to increasing road supply quickly in India (mostly political and administrative, not financial), compared to many other countries and China. Hence, there should be a far greater emphasis here on using roads efficiently, and far less talk of innovative ways of financing them through Special Purpose Vehicles (SPVs).

By using road infrastructure efficiently, I mean optimising on the amount of road space used by a passenger""in terms of the Road Area Per Passenger Kilometre (RAPP) travelled. It should be noted that merely measuring road area per capita in a city or country is not enough, without taking into account the amount of travel going on. A dirt poor country is one with very little roads or travel and hence road area per capita will be close to zero. As an economy gets richer, road area per capita will and should rise, but RAPP need not rise. And with well designed policies it can even fall, that is, road space efficiency can go up.

Data do not exist, but on specific road stretches, RAPP can be measured by dividing the total vehicle area by the number of passengers on that stretch, say a kilometre. For instance, if 100 people travel in 20 sq metre minibuses with 10 people each and another 100 travel alone in their sedans of 10 sq metre each, then RAPP, the average, is 6 sqm.

Optimising and reducing RAPP is not necessarily the same as providing or ensuring profitable infrastructure, which can be paid for by toll road and specific user charges. In practice, much infrastructure ends up being financed by direct or indirect, explicit or implicit taxes: all nontoll roads would fall under this category. Infrastructure financed by general taxes can be commercially viable, although not directly profitable. Even though taxes are better than deficit financing or inflation tax, they are still a questionable burden on the public, unlike toll roads paid for solely by users.

However, there is much more to road space efficiency than covering financial and/or private costs. Building roads and an associated rise in vehicles ignore social costs and social efficiency. More roads entail more emissions, chopping down trees, demolition of classic buildings and destruction of city heritage, and finally, displacement of villagers during land acquisition, sometimes even suicides. Charging carbon emission taxes alleviates only one of these costs. There are tradeoffs between building more roads and all the associated costs described above. Massive attempts need to be made to optimise on this tradeoff but very little has been done so far. Further, the financial costs of building roads, especially inside cities, are enormous.

It is blindingly obvious that instead of just trying to increase road supply, we should also try to use them efficiently. This is not the same as merely curbing demand which would lead to less travel, defined as passenger kilometres. More travel results from growth, and that is desirable, apart from unavoidable. However, policy makers need to start exploring direct measures of overall road efficiency, such as RAPP. Capital-output ratios are of no use for this purpose.

What I am recommending are policies to reduce RAPP. Such a reduction will provide many huge benefits, apart from saving on road construction and its various costs. RAPP can be reduced using a Vehicle Area (Annual) Levy, or VAAL, as I had suggested earlier ("Unclog Roads with a Vehicle Area Tax...Economic Times, Jan 6, 2006). Levy is a better and more accurate term than tax. VAAL should be strictly revenue neutral, that is, the government should not be allowed to use it to raise taxes. The revenues should be used to replace all other vehicle and road taxes and even petrol taxes, and any remaining surplus then used to offset other taxes.

The beneficial impact of VAAL on RAPP is obvious. Suppose everyone commuted to work alone in a 9 sqm sedan. After a stiff VAAL is imposed, suppose half trade down to going alone in 4 sqm minicars. The other half start commuting by 20 sqm minibuses that carry 10 people each. RAPP falls sharply to 3 sqm, a third of that earlier. Existing road area gets freed up, and the amount of new road area that needs to be built also falls sharply.

This trivial example does not address the difficulties imposed by VAAL, nor the phenomenally fierce resistance that it will encounter, even if strictly revenue neutral, as recommended here. Even if car owners do not mind commuting to work in a shared minibus, they would generally resist selling their cars that they wish to possess for non-commute driving and pleasure travel.

More crucially, for VAAL to "deliver", what are the chances of a commercial network of vehicles, such as minibuses, evolving to adequately provide transport for the public? Why not focus on developing public transport, and if needed impose congestion charges for peak time in busy areas like the central business district, as London did in 2003? VAAL is ideally suited to third world conditions, compared to all other alternatives.

The author can be contacted at vivekmoorthy@iimb.ernet.in