In last week's post, I began to introduce Lucknow's real estate market in an attempt to unearth the story behind the city's residential segregation along religious lines. I have shown where Muslims achieve higher and where lower prices than non-Muslims for comparable property (importantly, as one commentator pointed out by email, adopting a seller's perspective -- for sellers, higher prices are good, while for buyers, lower ones would be. I am currently wrapping my head around how this impacts my findings). Today, I want to volunteer one explanation for why this price difference (still firmly from a sellers perspective) might be as it is: it has to do with social networks and proximity - and with the uneven opportunities in colluding with the "actually existing state".

To begin, however, we need to cast a more critical look at my raw data source, the city's Property Index Register. In my last post, I have taken the same at face value, assuming that it reflects the real economy in a straightforward way. While the general spatial distribution of house prices seems to be about right, however, this assumption is significantly flawed. The property index register is kept in order to levy stamp duty on the market value of sold houses, an endeavour frought with corruption - nobody likes to pay more taxes than necessary, and at least some bureaucrats will certainly be open to negotiation in exchange for a bribe. In fact, the Uttar Pradesh Stamp (Valuation of Property) Rules of 1997 provide for a maximum fine of a mere 500 Rs should anybody furnish wrong information - this is almost an invitation! I think it is fair to say, therefore, that whatever the property index register records is the outcome of a negotiation process with the officers involved - rather than (or at least on top of) an uncomplicated matter of supply and demand.

These officers, however, are embedded in social relations and have identities of their own, casting the "actually existing state" 1 in a much more social light than political theory usually acknowledges. Most officers (and indeed all SROs in Lucknow, as far as I know) are non-Muslims, and this is where it becomes interesting: what if Muslims have lower chances of colluding with them? After all, corruption is a much more complicated endeavour than a mere economic exchange of bribes. How woudl this affect my analysis?

Since stamp duty is levied on assessed market value rather than actual sales price, undervaluation of property - the chief route collusion takes - enters the regression presented last week through the control variable "approximate property size" (which is market value divided by applicable circle rate). The regression has shown that, in new Lucknow, Muslims achieve lower prices for comparable properties. If, however, they also have less opportunities to collude with non-Muslim bureaucrats and thus only "negotiate" a less undervalued property assessment than non-Muslims, the regression would compare apples with pears: it would only show that Muslims achieve lower prices for a property of lower real value (rather than for a property of similar real value). In other worsd: if we assume that Muslims are less successful in collusion, their apparent discrimination through market prices vanishes. But are they less successful?

Surprisingly, even this can be shown by data - if we look at stamp duty levied (yep, you heard me right: evidence of corruption and state discrimination in official databases - I love my India!). A second regression indeed shows that Muslims pay higher stamp duty on properties of the same approximate size, sold at the same price, in the same mohalla, on the same date than non-Muslims - in most parts of Lucknow (again: if you want the full statistical details, drop me a line). The statistical effect is never large, but always significant. The following map breaks it down by ward (larger version):

Now does it matter whether Muslims are discriminated against by the market or the state, you might ask? It arguably does: the former would suggest popular anti-Muslim sentiment, while the latter shows the social life and political economy of bureaucratic collusion. I am not yet sure if my evidence is strong enough to say that it can explain residential segregation, heck, I am not even sure I got my economics right here. More work has to be done and more thoughts spent on this - for which I hope for your input. What do you think about Muslims' role in real estate? What are your experiences with corruption? And how far would you trust the analysis I presented?