LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI

Shirish B Patel

To be published shortly in Economic and Political Weekly and in the April 2013 issue of Environment and Urbanization 27 / 403 Abstract: Many Indian cities have evolved over centuries, through the Middle Ages. In The World Bank has been general our cities do not have the grand relentlessly complaining that Indian layouts and leafy avenues of many of cities are not optimally using their land. their more recent Western counterparts. They have been continuously pressing New Delhi is an exception. So are for upward revisions of FSI, particularly the cantonment areas that are an in Mumbai and Bengaluru. extension of many of our older towns. But then these layouts are a Western However, the bald comparison imposition and have nothing to do with of FSI (Floor Space Index) across cities the way old Delhi or other older cities that the Bank presents is seriously developed. Bengaluru was laid out misleading. This paper is about the more or less like a cantonment town. other factors that affect the way a Mumbai is a mixture of very different city functions. The author proposes layouts, from some parts that are the a new metric, Crowding, defined as native old town and others like the Fort the number of persons per hectare for area and Dadar that are differently and a particular urban use. Thus we have more spaciously planned. What should Indoor Crowding, Park Crowding, be obvious, but apparently is not to the Amenity Crowding, and crowding city’s administrators, is that you cannot of any other activity we consider have the same single set of uniform significant. But the most useful of all building and development rules applied these is Street Crowding, and a most to every part of the city, without regard intriguing relationship emerges between to the differentiation called for by such Indoor Crowding, Street Crowding, varying urban layouts. the Plot Factor which characterises the urban layout, and FSI. The choice of In every city buildings conform to a set of appropriate FSI thus has to be in the building codes—informal earlier, now context of these other three parameters. formalized. These codes are complex, It cannot, as the Bank suggests, be and usually vary from one locality to independently and arbitrarily set. another: buildings have to observe a minimum required front open space, In the second part of the paper possibly side and rear open spaces the author indicates how the new as well, with plinths often mandated to metrics of Crowding might be used in cover no more than a specified fraction planning or re-planning urban areas of the plot area. The maximum number in general. of floors is also often specified. These older, stronger regulations of urban form give us precincts like Marine Drive, where all buildings have a

404 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI common front line, and an identical more importantly in regard to transport footprint; or Ballard Estate, where front and the crowding on the streets. The FSI lines are similarly required to be rigidly also should be allowed to be different adhered to. The most extreme example in different parts of the city, depending of rigorous building control is perhaps on each locality’s layout. Note that a Paris, where buildings touch each other layout is hard to change, in regard to on either side, but in many localities all streets and plot boundaries, once it has owners on a particular street front have been initially set and construction has to adhere to a common façade line and taken place. identical floor levels, and quite possibly the material of façade construction. There can be no doubt that the more compact a city the more efficiently A post-World War II innovation from it will function. Networks of services Americaii,iii,iv introduced a new form will be smaller. Travel distances will of building control. This is called FSI be reduced. And of course a certain (Floor Space Index) in , FSR (Floor minimum amount of FSI is needed to Space Ratio) in a few countries, and produce enough presence to make life FAR (Floor Area Ratio) everywhere else on the streets interesting and stimulating. in the world. It is the ratio of built-up Nobody can want large derelict areas area of all floors on a plot to the area within the city. So it is clear that FSI in a of the plot itself. The FSI regulation is locality should not be too low. generally welcomed by architects. They like the freedom to reduce the footprint But the World Bank is at the other of the building, shape it or distort it as extreme. It has been relentlessly they please, and increase the number complaining that Indian cities are not of floors, while still observing the FSI optimally using their land. The Bank rule which sets the total built-up area particularly picks Mumbai to make allowable on each particular plot. From this case and points out that the city’s the authorities’ point of view, setting irrational building rules impede good up building regulations is now greatly economic use of real estate. Alain simplified. Setbacks can be made Bertaud in particular, consultant to the a function of the proposed building World Bank, is insistent that the Floor heights, façade lines or floor levels no Space Index (FSI) levels in Mumbai are longer need to be worried about. Only too low and need to be immediately the FSI specified has to be carefully and substantially increased.v,vi Once managed to ensure that the extent this happens, it will become a model of built-up floor space permitted in a for the rest of the country. To underline locality does not exceed that locality’s its argument, the World Bank presents infrastructure capacity, in regard to a bald comparison of FSI across water supply and sewerage of course but international cities, which is both

27 / 405 meaningless and misleading. It is Showing a table like this, without like comparing individuals’ weights mentioning in an adjoining column that without considering their heights or in Mumbai floor space consumed is 5 the societies they live in. The policy sqm (square metres) /capita, whereas recommendations that emerge can be in Manhattan it is 55 sqm/capita, is both dangerous and damaging for seriously misleading. If we were shown the city. these occupancy figures it would be clear that for the same head count of The following is a typical table from the people in a locality, FSI 1 in Mumbai is World Bank comparing FSI for several equivalent to FSI 11 in Manhattan. cities around the world: But it is on the strength of bald Table 1: Centre-City FAR values iii comparisons as shown in the Table above that Mumbai’s FSI has been portrayed as undesirably low. It has City FAR already been pushed up in many cases Sao Paolo 1 to 4, which is the current upper limit, Mumbai 1.33 except in the case of hotels, educational institutions, hospitals and the like where Chennai 1.5 the limit can be much higher; and for Delhi 1.2-3.5 redevelopment of tenanted properties, where there is no upper limit. Amsterdam 1.9 Venice 2.4 However, there are three major Paris 3 factors missing in this World Bank’s comparison of FSI. The first is that cities Shanghai 8 are at different levels of economic Vancouver 9 development, inhabited by individuals San Francisco 9 occupying, on average, different extents of floor space. Living is simply Chicago 12 more crowded in some parts of the 12 world, and less crowded in others. Here is a comparison of what we might Los Angeles 13 call Indoor Crowding in Manhattan’s New York 15 CD-8, more commonly known as the Denver 17 Upper East Side, and Mumbai’s C Ward (each reputed to be the most Tokyo 20 street-crowded residential district in 12-25 its city):viii, ix Source : World Bank (2012)

406 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI Table 2: Indoor Crowding in CD-8 and C Ward

Persons / hectare of built-up Persons / hectare of built-up floor area Locality floor area (commercial) (residential) Manhattan CD-8 (Upper 157 (or 64 sqm per person) 235 (or 43 sqm per person) East Side)

Mumbai C Ward 1,014 (or 10 sqm per person) 1,186 (or 8 sqm per person)

The table highlights staggering • SC = Street Crowding = Occupants / differences: Manhattan’s CD-8 has hectare of Street Area more than 6 times as much residential floor space per person, and over 5 • IC = Indoor Crowding = Occupants / times as much floor space per job, as hectare of Built-up Area compared to someone in Mumbai’s C Ward. With such extravagant use, no • PF = Plot Factor = Buildable Plot Area wonder Manhattan needs a so much / Street Area larger FSI. • FSI = Built-up Area / Buildable The second factor that is missing in Plot Area comparing localities is the extent of buildable area: in other words, the The following is their relationship: proportion of buildable plots to street area. We call this the Plot Factor, and it SC = IC x PF x FSI X matters because FSI applies only to the area of buildable plots. Street Crowding is thus the product of Indoor Crowding (the number of And finally, what must be factored into occupants per hectare of floor area), any debate on FSI is Street Crowding. the Plot Factor and FSI. The suggestion here is that we introduce this as a new and important index to The analysis of crowding in a locality give us an idea of how many occupants can and should be further refined. For live within a given street area, and example, if we say: consequently how crowded the streets will be. HC = Home Crowding = Residents / hectare of Home Built-up Area In any locality, if we adopt the following definitions: JC = Job Crowding = Job holders / hectare of Jobs Built-up Area

27 / 407 SC = Street Crowding = Occupants / In calculating the Street Area it is hectare of Street Area important that we exclude the areas devoted to parking, and the areas PC = Park Crowding = Occupants / devoted to arterial traffic. Just as we hectare of Park Area exclude surface railway areas when computing street areas, we should also

PFH = Plot Factor (Homes) = Home Plot exclude that part of each street that is Area / Street Area occupied by arterial traffic, or taken up by a particular function like parking.

PFJ = Plot Factor (Jobs) = Jobs Plot Area / Street Area Moreover, it is a fact that different societies use their streets in different xi FSIH = Floor Space Index (Homes) ways . When looking at Street = Built-up Area (Homes) / Plot Area Crowding in different localities we (Residential) should no doubt apply an appropriate multiplier, or attenuator, to particular

FSIJ = Floor Space Index (Jobs) = Built- countries’ values before making up Area (Jobs) / Plot Area (Commercial) comparisons internationally. And we should not forget the value of transition Then we have: open spaces, which are between those spaces that are entirely private and

SCH = HC x PFH x FSIH (night-time) those spaces that are entirely public. Transition Space Crowding deserves

SCJ = JC x PFJ x FSIJ (daytime) to be a valuable and separate metric from Street Crowding, if for no other We are not saying that every resident reason than to recognize the value of will be out on the streets at the same transition spaces.xii time, or that all job holders will similarly be out on the streets simultaneously The following is a comparison of the during the day. But some fraction of distribution of Land Use in Manhattan each set will be out on the streets at any and in Mumbai’s Island City: given time (with a maximum reached perhaps during the evening rush hour). And the point of our exercise is not to determine that precise number, but to assess the impact on Street Crowding of changes in FSI.

408 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI Table 3: Land Use in Manhattan and Mumbai

Land Use Manhattan (2000) Mumbai Island City (2005)

Streets 26.4% 28.6%

Homes 25.0% 24.5%

Office 13.6% 19.4%

Industrial 1.6% 4.9%

Amenities 9.3% 2.9%

Transport 5.1% 6.4%

Open Space 14.2% 3.8%

Rest 4.8% 9.3%

Total 100.0% 100.0%

Street Areas as a proportion of total Amenities and Open Spaces, which land area in Mumbai’s Island City adds up to 23.5% in Manhattan and and Manhattan are not very different. 6.7% in the Island City. This difference Land for work areas is roughly 40% is more glaring when seen as Amenity in Manhattan and 49% in Mumbai. Crowding and Open Space Crowding: The startling difference is in land for

Table 4: Crowding in Manhattan and Mumbai

Manhattan (2000) Mumbai Island City (2005)

Residents 1,529,357 3,086,362

Jobs 2,085,315 1,974,680

Area ha. 5,775 6,882

Plot Factor 1.52 1.71

Total Factor 3.79 3.49

Transport 5.1% 6.4% Crowding persons / hectare Street (night) 1,003 1,567

Street (day) 1,768 1,630

Amenity 2,841 15,223

Open Space 1,870 11,743

27 / 409 Amenity Crowding in Mumbai is The following is a crude comparison of more than 5 times higher, and Open C Ward in Mumbai and Manhattan’s Space Crowding 6 times higher than CD-8, with no correction for parking or in Manhattan. for arterial traffic:

Table 5: FSI and Street Crowding in CD-8 and C Ward

Locality Street Crowding persons / hectare Residential FSI

Manhattan CD-8 2,190 6.26

Mumbai C Ward 4,690 2.26

Despite having an FSI that is about one- by the slow rise in car ownership, and third of Manhattan’s Upper East Side, was much more closely related to the Mumbai’s C Ward‘s streets are more increase in building volumes. than twice as crowded. Any doubling or quadrupling of FSI in Mumbai’s C The suggestion here is not that there Ward, as is being currently debated should be no change in Mumbai’s FSI under prodding by the World Bank, regime. Where existing residents would will only amplify its already massive simply like to have more floor space, street crowding to levels so far not with no increase in the number of seen anywhere in the world; not to families in a locality, there is surely no mention further aggravation of the harm in raising FSI. In the older parts already excessive Amenity and Open of Mumbai the existing, already built- Space Crowding. up FSI levels exceed 2 and sometimes 3, but reconstruction is permitted only We have recently witnessed in a small to the more recently set 1.33. This is way in Mumbai’s suburbs the impact absurd. And there is no reason why FSI of lifting FSI. In October 1997, there should not be higher around new transit was a sudden relaxation and FSI nodes, where added population would could be increased from 1 to 2 in the place no higher burden on existing Western suburbs, using Transferable transport capacities; provided of Development Rights. Everyone in Bandra course that water supply and sewerage experienced the sudden increase in are properly taken care of, as well as traffic volumes, a new and infuriating Amenity and Open Space Crowding; congestion, arising from a doubling of and provided also that Traffic Demand traffic, which could not be explained Management (TDM) policies are properly implemented.

410 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI But the World Bank is focusing in a worsening of living conditions, exclusively on FSI, disregarding all other particularly travelling conditions. aspects of urban life. Its comparison of But worst of all, it is a red herring. It international FSI values is simplistic in distracts us from the central problem. the extreme and seriously misleading, This is that adding to the city’s land area because it ignores the other equally by establishing new transport arteries relevant parameters of Street Crowding, is being invariably and unaccountably Indoor Crowding and the Plot Factor, delayed. The high cost transport apart from Amenity Crowding and projects being given the highest priority, Open Space Crowding. Pressing like the Western Coastal Road, or the for a major upward revision of FSI Virar-Churchgate railway, offer to open without a corresponding improvement up no new land for the city. There is in infrastructure, particularly transport also a deliberate companion policy of to deal with street crowding, is withholding land from the market on logically indefensible. Because it can one pretext or another. How to keep only aggravate every type of public land in short supply is thus the name of crowding, it promises something it the game, by which windfall fortunes cannot deliver, an improvement in the can be made in short order, and never quality of life. For the poor, and the mind what happens to Mumbai or vast majority of citizens, it will result its citizens.

II

When we convert the foregoing Total Factor (TF) = Total Area / formula into a graph we get some Street Area interesting new insights. And to arrive at Gross Densities, we introduce a new And we notice that Gross Density (GD) parameter, Total Factor (TF) defined is a function of Street Crowding (SC) as follows: and Total Factor (TF):

GD = SC / TF

27 / 411 Fig. 1: Localities seen on the Indoor / Street Crowding Graph

Indoor Crowding is represented as our locality’s FSI value, we turn left persons / hectare on the positive X-axis. again and move downwards, where Street Crowding is on the negative on the negative X-axis we read off X-axis. The diagonal radial lines in the Street Crowding, the product of IC, PF first quadrant are the multiplier lines for and FSI. Plot Factor. If we start with any given level of Indoor Crowding (IC), and In the third quadrant we introduce move up to the relevant Plot Factor diagonal lines representing various (PF), then turning left at that point and values of the Total Factor. Continuing extending the horizontal line leftwards our vertical line downwards from the brings us to the product of IC and Street Crowding value, until it reaches PF on the positive Y-axis. The second our locality’s value of Total Factor, and quadrant contains the diagonal radial turning again we move horizontally lines for FSI. These are again multiplier rightwards to read off the value of lines. Once our horizontal line reaches Gross Density on the negative Y-axis.

412 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI The graph shows the diagram for which could be added in without each locality. The values for various much altering the parameters relating parameters are given in Annex 1. to Plot Factor, FSI, and Indoor or Street Crowding. Mumbai’s C-Ward is mostly native town, but to the west has a reclaimed Notice the following: extension that is well laid out with large green spaces. D-East is a mixed • Values for Indoor Crowding in residential and commercial district, Manhattan are remarkably low, both properly laid out and earlier occupied for CD-5 and CD-8; and are much by a mixture of British and native lower than for any locality in Mumbai; inhabitants. D-West is the most pricey and upmarket residential district in • Street crowding in CD-8 is also lower all of Mumbai, with large vacant than for any locality in Mumbai; tracts occupied by the Governor’s establishment and a wooded area • In Mumbai, the horizontal line that holds the Parsi Towers of Silence seems to rise with diminishing (the area for disposal of their dead). prosperity—the poorer the locality G-North (South-East) is partly properly the larger its rectangle in the laid out, and partly Dharavi. upper quadrants.

All these localities, and Manhattan’s • The Gross Density for residents is CD-5 and CD-8, are over 200 hectares nowhere higher than 1,000 persons in area. We have chosen this size so / hectare in Mumbai. This is despite that each such locality can be expected not having enough land set aside to have within it amenities and open for amenities and open spaces. A spaces sufficient at the very least for its figure of 750 appears more realistic own needs, if not additionally offering as an absolute upper limit for global facilities for the rest of the city. planning, as exemplified by Charkop which does have schools and open The exception is Charkop, which is only spaces. And the figure of Gross Density 57 hectares in extent. It is shown here should be calculated on the residual on the diagram because it has schools area after deducting what is needed for and parks, and could be replicated arterial transport. without much change to several times its present area. It is also an admirable Seeing these localities on a diagram development: well integrated mixed- throws up two interesting questions. The income housing incorporating sites- first arises from the observation that on and-services for low-income groups. a greenfield site, or a brownfield site The only activity missing is employment, undergoing major redevelopment, we

27 / 413 can have any particular level of Indoor words, with IC and SC determined, Crowding, and any particular level of we can still juggle PF and FSI, as long Street Crowding, and still have a choice as their product remains unchanged. of how high or low the horizontal The following diagram illustrates the connecting line should be. In other possibilities :

Fig 2: Varying PF and FSI

In the diagrams we have assumed corresponding Total Factors of 5, 4 and Indoor Crowding at 1,000 persons / 3 for each of the three cases. Thus we hectare, and Street Crowding at 3,000 see that in these three cases, as the persons / hectare. For three possible horizontal line descends, the Gross values of Plot Factor, at 3, 2 and 1, Density increases from 600 to 750 to we get FSI values of 1, 1.5 and 3 1,000 per hectare. respectively (the product, at 3, is the same in all three cases). Let us assume We should note that what changes in that in each case the area for Amenities these cases, with gross density rising and Open Spaces is equal to the Street from 600 to 750 to 1,000, is the Area: in this way the Amenity and character of the development: with a Open Spaces Crowding will remain Plot Factor of 3 and an FSI of 1 we unchanged in all three situations, and will have low-rise buildings, probably will be the same as the value for Street no higher than 2 or 3 floors; with FSI Crowding. With Plot Factors at 3, 2 1.5 the number of floors will go up to and 1 we have streets taking up 20%, 4 or 5; and with FSI 3 the buildings 25% and 33% of the total land area, will be 8-10 stories high (assuming and Amenities and Open Spaces doing footprints are about a third of the plot). the same. Plot areas will be 60%, The texture of the locality changes, and 50% and 33% respectively. So we get

414 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI so does the quality of life on the street, The second interesting question that can despite Indoor and Street Crowding be studied on the diagram is the effect numbers remaining identical.xiii of a sudden rise in FSI in any locality. Let us imagine that we have a locality with In a city we probably need a mix of Indoor Crowding of 1,000 persons / all three intensities of development: built hectare, and Street Crowding of the denser developments near the 2,000 per street hectare, a Plot Factor major transit nodes, and the lower of 2 and FSI of 1. The FSI is suddenly density with its shallower-rise buildings raised to 2. further away.

Fig 3. Effect of sudden increase in FSI

We have an initial situation ‘A’. The the area does not change: they simply Plot Factor is fixed, with plots and move into the larger accommodation streets already defined. Now we allow provided by the increased FSI. The the sudden increase in FSI from 1 to 2, top right corner of the initial diagram as happened in the Western Suburbs ‘A’ descends along the PF diagonal to in the late 1990s. The first possibility a new location ‘B’ which reflects the is that the number of families living in now improved level of reduced Indoor

27 / 415 Crowding. Street Crowding remains The horizontal line ‘C’ will then extend unchanged. This is the argument put leftwards till it reaches the new FSI, forward by Alain Bertaud, who sees and at this point descend to a new and increased FSI only as something that higher level of Street Crowding, and benefits existing families, with no higher Gross Densities. negative consequences. The reality will most likely be somewhere The second possibility is that existing between these two extremes. It certainly families cannot afford the extra cannot be the case that all the new accommodation provided by the floor space will be taken up only by additional FSI, and that new families existing families and that despite a move into the extra floor area. In the sudden change in FSI there will be worst case these new families will no worsening of Street Crowding occupy accommodation no larger than or Amenity Crowding or Open what is occupied by the earlier families. Space Crowding.

CONCLUSION

What has been presented above placing heavier demands on resources, is a proposal for a new metric we particularly urban transport. But above could call ‘Crowding’. This should be all, FSI decisions should be taken not useful in measuring the characteristics randomly but in the context of each of existing localities in regard to a particular layout and the desired levels variety of aspects of urban life. Thus of indoor and outdoor crowding. we can consider Home Crowding, Job Crowding, Amenity Crowding, Open Space Crowding, Transition Space Crowding and Street Crowding. After comparing a number of localities around the world hopefully we can develop norms that set out the desirable range of values for each of these aspects of urban crowding. We do not want values that are too high, nor should we want values that are too low. Too high would imply uncomfortable living; too low would imply derelict or lifeless areas that are uncomfortable in a different way, apart from probably

416 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI Annex 1

Table 6: Localities in Manhattan and Mumbai comparedxiv,xv

All Users All Users Total Factor All All Users Area Indoor All Users Street = Gross Users Locality Population Jobs Plot ha. Crowd- FSI Crowd- Area / Street Gross Factor ing ing Area Density

Manhattan

CD-5(Midtown) 424 44,028 894,290 296 16.04 1.26 5,986* 2.70 2,216

CD-8 (Upper 513 217,063 137,645 180 7.29 1.67 2,190* 3.17 692 East Side)

Mumbai

A-Mid 345 53,735 243,681 638 3.66 0.73 1,695 1.97 174

B 279 110,415 31,349 692 1.79 0.88 1,086 2.14 508

C 214 202,686 102,122 1,018 2.05 2.39 4,992 3.50 1,425

D-East 216 163,627 149,127 1,201 1.75 2.46 5,182 3.58 1,446

D-West 262 83,462 15,836 366 1.34 4.71 2,317 6.11 379

G-N/S-E# 314 299,897 228,607 3,116 1.13 1.04 3,682 2.19 1,682

Charkop 57

39,466 1,058

1,058 1,157 1.02 2.96 3,488 4.89 713

Island City 6,882 3,086,362 1,974,680 1,042 1.12 2.20 2,570 3.49 735

Notes * : Manhattan has an underground railway # : Includes Dharavi

27 / 417 i “Life Between Buildings” is the title of a wonderful book written in x The right hand side of the equation when expanded reads: (Occupants the 1970s by Jan Gehl, architect and Professor of Urban Design in / Built-up area) x (Buildable Plot area / Street Area) x (Built-up area / Copenhagen. He distinguishes between essential activities, optional Buildable Plot area). The Built-up area and Buildable Plot area terms activities and social activities that take place in the public space cancel out, and we are left with Occupants / Street Area, which is the between buildings. His focus is on the design of public spaces definition of Street Crowding. to maximise social activities. Ours in this paper is rather on how indiscriminate increases in FSI can make even essential activities xi A friend to whom this paper was sent for comments and suggestions, unacceptably difficult. says: “Perhaps you need also to draw attention to what crowds our streets: people, motorised and non motorised vehicles of all sorts ii Gaspers, David A., “Form-Based Code as a Regulatory Tool for Mixed (two-wheelers, three-wheelers, four wheelers from cars of different sizes Use Urban Infill Development in Lincoln, Nebraska”, 2006. to vans, trucks, buses, coaches), lack of discipline and road manners on the part of drivers, street parking, handcarts, bullock carts, hawkers, iii Sussna, Stephen, “Bulk Control and : The New York City shops spilling over onto the street, rubbish dumps, utility installations Experience”, Land Economics, Vol. 43, No. 2 (May, 1967), pp. 158-171, (e.g. transformers), street furniture, trees, shacks of the homeless, (http://www.jstor.org/stable/3145239). stray animals, community dogs, protest groups, badly paved pedestrian pavements, ratio of vehicular road to pedestrian road, etc. There is one iv Gallion, Arthur B. and Eisner, Simon, “Urban Pattern”, D van hell of a difference between a New York or Washington Road and a Nostrand, 1965. Bombay Road based on the uses to which it is put.” The implicit suggestion that street functions would benefit by being v Bertaud, Alain, “Mumbai FAR/FSI conundrum. The perfect storm: The segregated is something that Jan Gehl for one would strongly resist. four factors restricting the construction of new floor space in Mumbai.” When pedestrians and traffic are completely separated he thinks it (available at http://alainbertaud.com) , 2011, mentions Mumbai’s becomes duller to drive and duller to walk and duller to live along estimated average of about 4.5 m2 per person in 2009. these streets. But he likes the idea of stopping traffic at the perimeter of a locality, with a pedestrians-only area in between, and a short vi Bertaud, Alain, and Brueckner, Jan K., “Analyzing Building Height walk to your particular door or gate. Hermann Knoflacher, an Austrian Restrictions: Predicted Impacts, Welfare Costs and a Case Study of transportation engineer, echoes this when he suggests that the distance Bangalore, India”, World Bank Policy Research Working Paper 3290, from where you live, or work, to your car should be at least as much as April 2004. the distance to the nearest bus stop.

vii Brueckner, Jan K. and Shridhar, Kala Seetharam, “Measuring xii Between the private indoor space and the public street there often Welfare Gains from Relaxation of Land-Use Restrictions: The Case is—and it is certainly desirable—a transition open space, which is of India’s Building-Height Limits”, (available at http://www.socsci.uci. neither entirely private nor entirely public. These are common spaces edu/~jkbrueck/india.pdf), 2012. not closed to the public but nor are they casually wandered into by the public. The size and configuration of these spaces matters and affects viii The author first came across the term “Crowding” recently, in a paper the experience of moving from indoors to street. Courtyards work well. by Rémy Prud’homme, “Seven Notes on Mumbai’s Growth and How to And courtyard crowding, like other types of crowding, needs to be taken Finance it”, May 2007. But Prud’homme uses it only to express Indoor into account. Crowding, and does not extend it, as proposed in this paper, to other aspects of urban life. xiv Data for Manhattan kindly provided by Shampa Chanda, Director of Citywide Planning, Department of Housing Preservation and ix The author gratefully acknowledges the value of fruitful discussions Development's (HPD’s) Planning and Pipeline Development Division. and research support from Prof Abhay Pethe and his post-graduate students Sahil Gandhi and Vaidehi Tandel at the Department of xv Data for Mumbai from Biond, a Mumbai-based GIS company. Economics, University of Mumbai; and Vidyadhar K Phatak.

418 / 27 LIFE BETWEEN BUILDINGS : THE USE AND ABUSE OF FSI