Improvising Sanitation
When sanitation provisions are absent or unreliable, people are forced to improvise solutions, however temporary. Improvisation is not only a spontaneous act. Instead, it emerges from the routines and rhythms of everyday life and constitutes a strategy for coping with uncertainty, disruption and breakdown. It is a learning process developed over time, but one deeply fractured by income, contacts, ethnicity, and gender.
In Rafinagar Part-2, at the edge of the large stormwater drain (nallah) that runs on the west side of the settlement, residents have built collective makeshift toilets out of cloth, timber, jute and iron sheets bought, found or salvaged from waste. Each toilet is used by the 15-25 households living in each lane. In some cases, where people have the time and skill, or find they cannot donate money to the construction but can donate their labour, they have built these makeshift toilets themselves. In other instances, where the residents of the lane have been able to raise enough money between them, they have employed someone to build the toilet. In these lanes, each family typically contributes between Rs.100-200 towards construction (a total of Rs.3000-5000). These toilets are often referred to as 'hanging latrines' in the sanitation policy literature, and given the untreated nature of the sites they are located in they are often areas where people are highly vulnerable to illness and disease.
Although we are calling these makeshift toilets, they are not necessarily temporary infrastructures. The first makeshift toilets were built in Rafinagar Part-2 almost eight years ago, and in many cases they have become a permanent sanitation solution: an improvised toilet becomes routinised. Indeed, when they are damaged by the high and low tides in the large stormwater drain which not only washes away the waste but also gradually rots, damages, and washes away the structure or parts of it, residents once again contribute their money and/or time and/or labour to repairing or reconstructing the toilet. In some cases the toilet is rebuilt once per year. In other cases, demolitions carried out in Rafinagar Part-2 by BMC bulldozers have loosened the earth that keep these structures stable, destroying not only people's houses but also the toilets built over the nearby stormwater drain. This demolition does not have a predictable rhythm: some parts of Rafinagar Part-2 were subjected to demolitions as recently as two years ago, while other parts have not been demolished for several years now. Residents in the lanes in these latter areas no longer feel an acute threat of demolition and have built more stable and lasting toilet structures, sometimes building two cubicles adjacent to each other so that men and women of the lane could have their own toilet. The investments that residents have made in the makeshift toilets are linked not only to the resources they have but to the sense of impending demolition they feel.

Makeshift toilets in Rafinagar Part 2.
Urban improvisation is an uncertain process. It can depend on the availability and condition of the materials, on people's capacities to intervene in infrastructures, on the maintenance and investment (or lack of it) by the state, and/or - as Sameera indicated - on the relations between residents:
"[The municipality] should come to clean the gutters. Unsanitary conditions are not nice to see. The residents here don‟t pick anything up. And there are fights. No one pays any attention to the gutters. They let it be. We face difficulties because near the [water] drum, there is a grate (jaali) so [waste] accumulates there… The one who lives on the other side [of the lane] has put it so that the waste doesn't go to their side. But it creates difficulties for us because the waste accumulates near our house."
Sameera's complaints about the family who lived a few houses downstream of the drain, on the other side of the lane, emerged because while this family had put in a grate, no one upstream of the drain had put in grates. As a result, waste came down and collected in the drain in front of Sameera's house. On the other hand, in many other lanes, the introduction of grates in the open drains had successfully helped to divide the responsibility of keeping the drains clean amongst households along the lane. Here, the two households living across from each other in the lane took responsibility for keeping the section of the open drain in front of their house - demarcated by a grate at each end - clean of solid wastes. This reflects how the collective nature of infrastructure renders both opportunities for and threats to successful improvisation, where the manipulation of infrastructure depends on a certain collectivised becoming of infrastructure.

The use of grates in the open drains constructed in lanes can help to manage the cleaning of the drains by households, Rafinagar Part 1.
In other parts of Rafinagar, although the drainage infrastructure has been upgraded to closed drains, not all residents perceived these upgraded drains to be better. This is because the BMC does not clean these drains in Rafinagar and they are more difficult to improvise with compared to open drains. Residents must be able/willing to put in more labour (to lift the heavy drain covers and clean the drains) or money (to employ someone to do so) in order to maintain these upgraded infrastructures and make them work for them.

Closed drains built in some of Rafinagar's lanes. These offer very different possibilities for improvisation.
While improvisatory sanitation structures are often seen as last resorts, we would argue that policy can learn from and should seek to support the practices and responses that residents have already established over time.
