Relaxing FSI

With reference to Jan K Brueckner and Kala S Sridhar’s “In Defence of Relaxed FSI Limits” (EPW, 28 September 2013), my problem with arguments about whether or not to relax FSI (floor space index) is that it continues to be regarded as an independent variable, something that can be compared across cities around the world, without reference to the floor space families occupy. It is like comparing liver sizes without reference to body weight or drinking habits. In Mumbai a typical middle-class family occupies 5 sqm per capita. In Manhattan it occupies 55 sqm per capita. So FSI 1 in Mumbai would produce the same density of people on the ground as FSI 11 in Manhattan. Similarly, prescribing FSI 4 for poor families in Mumbai (as the Maharashtra government is currently doing) is ruinous in terms of overcrowding; it should be 4 in posh areas like Malabar Hill where families occupy large floor spaces. FSI relaxations are often desirable, but they need to be calibrated according to context, and FSI is not a universal good, of which necessarily the more the better.

The monocentric city model has its uses, but its fundamental limitation arises from the fact that it assumes only two uses for land: residential areas and radial streets for automobile transport to the single central employment location. It ignores the land that needs to be set aside for schools and parks. If you measure gains accruing from increasing FSI towards the centre by calculating the value of reduced travel times, you should also set off against this the loss due to more crowded schools and parks as the city densifies; not to mention light and ventilation as the city densifies beyond all reason. To say that all increases in FSI are only gain, because of reduced travel times, despite other amenities being made scarcer, at times as in Mumbai to the point of vanishing altogether, is too obviously wrong to be worth discussing at all.