Imprecise estimates of slums in the 2011 census could affect welfare programmes for least privileged groups

The recently published census 2011 report on housing stock, amenities and assets in slums, the first of its kind in the country, reassuringly announced that the number of urban slums has declined and the percentage of households in slums has dropped from 23.5 (2001) to 17.4. On the face of it, this reduction appears to be a significant accomplishment and calls for praise. The government has concluded that cities are not as worse off as they were assumed to be since the new numbers reveal a lower percentage of slums. But the truth is far from it. Not only have the absolute numbers increased, but even the enumerated figures are inaccurate. It is a gross underestimate that can mislead and undermine social housing programmes.

The new figures show that 13.7 million urban households live in slums. Together, they accommodate about 68 million people. This is far less than the number projected by the Committee on Slum Statistics/Census constituted by the Ministry of Housing and Urban Poverty Alleviation in 2008. Their comprehensive report points out that the slum population would have reached 93 million in 2011. One may tend to dismiss such projections since the field level counts, such as the census, are considered far more reliable than statistical estimates. Though theoretically correct, in practice, the census does not truly count all slums.

Its beginnings

Slum enumeration commenced only during census 2001. It profiled demographic characteristics but not housing stock and amenities. Nevertheless, it was an important beginning. To its credit, it recognised different kinds of slums — those formally notified and those recognised but not notified by State governments and local bodies. It even went a step further in identifying housing clusters, with a minimum size of 60 to 70 households or 300 population, that were poorly built, congested, and without sanitary and drinking water facilities as slums. What it failed to do was to survey all urban centres irrespective of the population size. The enumeration was limited to places with more than 50,000 population. As a result, only 640 towns were surveyed. The results showed that 42.6 million people lived in slums, a low figure that did not match ground realities. After the Parliamentary Standing Committee on Urban and Rural development’s intervention in 2002, the survey was extended to an additional 1,321 towns that had a population of 20,000 to 50,000. The final number of those living in slums rose to 52.4 million. Even this did not add up much when compared to other estimates.

Large differences

While the Town and Country Planning Office, the technical wing of the Ministry of Urban Development, statistically estimated the figure to be 61.8 million, the Indian Agricultural Statistics Research Institute calculated 75.2 million and the U.N.-Habitat’s Slums of the World report indicated that in India, 158.4 million lived in slums. Such a large difference did not compel the Office of the Registrar General and Census Commissioner (ORGI) to change its approach in 2011. “Statistically estimated figures may not reflect the ground relativities,” was its defence.

The problem with the census count is twofold — one, the number of towns surveyed and two, and the more important factor, the number of slum enumerative blocks. There are 5,161 towns in total. In 2001, the towns surveyed were only a fraction. Second, the census considered only large clusters, 60 to 70 households or a population of 300, for purposes of slum enumeration and overlooked smaller clusters. In many cities, particularly medium and small-sized, the sizes of slums are not large. In bigger cities, the slums are fragmented and scattered, which are not taken into account either.

The Committee on Slum Census had anticipated that ORGI may again underestimate numbers in the 2011 enumeration and cautioned it to get its definition of slums right and widen its survey net. But this did not happen.

Old criteria

Though the census 2011 expanded the scope and covered 4,041 towns, it was still 1,120 short of the total number. Despite many agencies pointing out that slum clusters as small as 20 households must be included for counting, the census office remained fixed to old criteria. It seems to take a narrow and less useful position — convention and convenience are more important than facts. Even on the matter of convention, ORGI falls short of international practice.

Since 2000, when the U.N. adopted the millennium declaration and listed the development goals (MDG), countries realised the need to measure urban poverty accurately. It was important to monitor the progress of MDG, in this case, Goal 7-Target 11, which aims to improve the lives of at least 100 million slum dwellers by 2020. U.N. Habitat mediated a consensus on slum definition and insisted on any cut off figure such as household numbers. However, the Indian census office seems to persist with it.

The problem is not only with ORGI. Local governments too have failed. At end of the first phase of house listing in 2011, the census office admitted that many cities were reluctant to earmark new slum blocks for enumeration purpose. Instead, it demarcated the same blocks used in 2001 and only 2,543 of the 4,041 towns reported the presence of slums.

All those vested with the responsibility of counting slums would say that it is expensive to carry out an exhaustive survey. But imprecise estimates tend to short-circuit welfare programmes and the least privileged groups end up bearing the brunt. Underestimation of slum numbers produces an illusion of adequate progress. Agencies such as the Planning Commission and the various ministries that swear by census figures would tend to reduce subsidies, scale down allocation and may even want the state to withdraw from social housing schemes. All one could hope for is that the census 2011 figures on slums is treated only as a provisional figure and not a firm fact.

srivathsan.a@thehindu.co.in

More In: Comment | Opinion