Subsidy to nowhere

Last week's disaster has got a lot of people thinking of all kinds of urban issues, of which housing is one. Which reminded me of something I wrote last year for India Together; it reminded me that fixing things in this city, if we want to fix them, is not as simple as simply demolishing stuff and/or giving stuff away free.

Longish, but bear with me.

***
The arithmetic is actually quite simple. You can do it yourself. Divide 4 million people by five: that's 800,000, the number of families and so the number of houses that should have been built. Divide by five again: that's 160,000, the number of houses that must be built every year of the Government's five year term. Multiply by two: 320,000, the number of houses that should have been built in the two years that the Government had already spent in power.

Compare to the number of houses that had actually been built and turned over to families as of March 31, 1997: 1146. Yes, 320,000 versus 1146.

In case you had not guessed, these figures refer to the then Maharashtra Government's famous slum rehabilitation scheme. The promise to house 4 million slum dwellers of Mumbai was a major plank of the Shiv Sena-BJP alliance's election campaign in the 1995 Maharashtra Assembly elections. It was a laudable promise, and for this reason alone: it was a recognition that slum dwellers are citizens with rights, with definite and vital contributions to make to the urban economy.

In fact, no different from the rest of us.

And in fact, too, a July 1995 Government report that was the basis of the scheme, written by the senior bureaucrat Dinesh Afzalpurkar, actually put it in writing:

The slums and hutment-dwellers of unauthorized structures form an integral part of this vibrant metropolis. All of them undoubtedly have a share in the growth, status and prosperity of this great city.

Now given prevalent attitudes towards slum dwellers, official and otherwise, this quote from the Afzalpurkar report is remarkable by any standards. I don't know of any similar description of slum dwellers, ever. This recognition of their rights alone was a long step forward. Of course, such praise needs tempering because this was an election promise; one made with an eye firmly on votes in slum areas. Still, if translated into reality -- in this case, meaning real houses for real slum dwellers -- election promises are quite all right. This one certainly gave millions of people visions of owning legal, pucca homes, complete with water and electricity connections.

By now, you're wondering why I'm writing about something that happened almost ten years ago. One reason is that the idea behind this scheme -- the so-called "cross-subsidy" -- is still touted, in various forms, as an answer to the problem of housing the poor. And since it is, the reasoning behind it needs some examination.

In the cross-subsidy regime, a builder agrees to build a number of small flats for the poor, to be sold at subsidised rates. In return, the Government gives him various concessions and incentives to build larger flats to sell to the middle-class at market rates. The presumption is that the money he makes on the market-rate flats will pay for the subsidy on the others, and still leave him with enough to make all this attractive.

The Sena-BJP government added this wrinkle: they proposed to hand those lower-cost houses over to slum dwellers, not at subsidised rates, but for free. Thus was the cross-subsidy idea taken to its logical extreme. And looking at the numbers for an extreme like this is a good way to learn just what it is all about.

As that number 1146 shows quite plainly, the visions I spoke of above remained just that: visions. In two years, about one-three-hundredth of the number of houses the plan called for actually got built; and the scheme fell away into oblivion after that grand fiasco came to light in March 1997.

Why this failure?

Part of the answer to that comes from looking at some more arithmetic.

Calculations show that under the scheme, given the cost of housing at the time, each 1000 free homes would be paid for by about 560 houses built and sold to the middle-class. (The numbers have changed slightly over the years as the cost of real estate has changed, but this is essentially the equation). Therefore, the scheme rests on one premise: that there will be a market for that higher-income housing, at a price that will pay for its construction, the free housing and a profit as well.

Is this a reasonable premise?

Well, for one thing, the arithmetic I started this article with does not count the middle-class houses. If 160,000 houses must be built every year to be given away free, about 90,000 houses must also be built every year to be sold (remember the 560:1000 ratio). This makes for an annual total of 250,000 houses that had to be built during the Sena-BJP regime.

How many was the industry building at the time?

The Afzalpurkar report itself answers that question:

the present level [of construction of houses in the city] ... does not go beyond 35,000-40,000 tenements per year.

So to make a success of the slum scheme, contruction would have had to be increased by a factor of at least six. How was the industry going to be able to do this? The report was entirely silent on this point. This is no trivial concern, because that kind of increase in construction does not happen overnight. How do you step up supplies of raw material, transport, water, sanitation, not incrementally, but six-fold?

But let's even assume builders and an apathetic Government could have pulled off that spectacular feat. There was a far more insidious fallacy in the scheme that really doomed it. Like everything else, the price of real estate is founded on supply and demand. The annual supply of 35-40,000 units set the price of flats at a certain level, and that price made the whole subsidy idea attractive to begin with.

What would be the impact of 90,000 additional housing units a year on such prices, such a market? You don't have to be an expert to answer that: the market will crash, prices will fall. In fact, by 1997 those things were happening anyway, for other reasons. The then-Sena Chief Minister, Manohar Joshi, was already pointing a finger at the depressed real estate market, warning that it had ruined the prospects of the slum rehabilitation scheme.

As even this brief analysis shows, the very logic of the slum scheme is a page out of Catch-22. Bombay's eye-popping real estate prices makes the scheme conceivable. Implementing it makes it inconceivable.

This is why the first two years of the BJP-Sena regime saw that grand total of 1146 flats actually turned over to slum dwellers. Whichever way you look at it, the scheme was misconceived, miscalculated and entirely missed the mark. Therefore, it was also implemented in the only way it could have been: half-heartedly.

Now I don't mean this as an indictment of the Sena-BJP Government in particular. It's just that their promise to give those homes away free was an absurd election gimmick. Frustrated slum dwellers who voted for them in 1995 remembered that absurdity, in the 1999 and 2004 Assembly elections. At the same time, that promise is an efficient way to analyse the idea of this subsidy.

Still, there have been more successful examples of the cross-subsidy scheme. One, in the northern suburb of Goregaon, comprises about 6000 units which are being turned over to their new owners about now. It has worked for two reasons.

First, it asks lower-income group buyers to actually pay for their flats, though still at a subsidised rate. Not only does this greatly improve the arithmetic. It also avoids giving away housing for free, an always dubious idea on many counts.

Second, the flats were built not overnight, but at a pace that's more in step with the rest of Bombay's housing market. The problem here is that as a result, people have had to wait many years -- some as long as 20 years -- to get their flats. But at least the construction of these flats has not had a significant impact on the market, thus still allowing for a certain level of subsidy.

This, it seems to me, is the only way the subsidy scheme can work.

Then again, telling the truth about this idea is not exactly going to bring the votes pouring in. Not quite as sexy as announcing during an election that you will give away free housing to 4 million slum dwellers.

Which, in the end, is a good reminder of the most valuable nugget of wisdom all this arithmetical analysis turns up; it also is the other reason I wrote about these ten-year-old events: Never believe election promises. Never, and especially never if they include the word "free."