The politics of Indian census data

Author: Vikas Kumar, Azim Premji University

Indian governments spend enormous resources to collect data — including 12 billion and 22 billion rupees on decennial censuses in 2001 and 2011, respectively. Yet they appear reluctant to release it. The latest decennial census data on religion, for example, which were released on 25 August 2015, were collected almost half a decade ago in 2011.

 Kashmiri Muslim women walk with children on a deserted street during a curfew in Kashmir, 12 September 2015. The most recent data released on religion in India was from 2011. (Photo: AAP)

During the past 15 years, governments of both national parties have on more than one occasion deferred to political expediency on the question of releasing demographic data disaggregated by communities. In the process governments have contributed to the politicisation of statistics. The troubled past of the census data on religion reveals systemic problems insofar as the statistical wing of the government is insufficiently insulated from politics.

The National Democratic Alliance (NDA) government led by the Bharatiya Janata Party (BJP) that conducted the 2001 census did not release the religion data. The data were released after the 2004 elections by the United Progressive Alliance I (UPA-I).

The NDA’s reluctance can perhaps be explained by the BJP’s Hindu nationalist commitments. Given the relatively higher growth rate of the Muslim population, the increase in the population share of Muslims was inevitable. So, if the NDA had released the data it would have seemingly validated the Hindu right’s concerns about the Islamic demographic ‘threat’ and the BJP-led coalition government would have come under enormous pressure to suggest steps to address the ‘problem’. Later, when the UPA-I released the data, the BJP promptly published a collection of articles to expose the hypocrisy of secular parties that overlook the threat posed by the higher growth rate of Muslims.

More recently, weeks before the 2014 elections, the BJP attacked ‘anti-national’ UPA-II for ‘suppressing census figures [on religion]’ because it was ‘ashamed to admit its failure to take the Muslims out of deep poverty’. Yet the present BJP government failed to release the data until recently. The party seemed unsure about how it will be affected by the release of the data on religion ahead of elections in five major states with above national average Muslim population shares.

After the release of the data right wing ideologues have revived the debate about alleged ‘population jihad’ that is disturbing inter-community demographic balance. They have suggested that access to government welfare schemes should be made conditional upon family size and violation of two-child norm should be criminalised. Unsurprisingly, the media has unanimously questioned the release of data weeks ahead of assembly elections in the state of Bihar.

But it is not clear why the UPA-II did not release the 2011 census data on religion. The explanatory note to question eight of the 2011 Census Household Schedule clearly links the identification of scheduled castes to their religious affiliation. The latter is canvassed in question seven. This means that the caste and religion data have to be sorted simultaneously. Information about caste (and even tribe) from 2011 was available as early as 30 April 2013, a year ahead of the May 2014 parliamentary elections. Even the religion data were leaked selectively to the media in January 2015.

Since only stripped down excel tables on major religions were released in August 2015, the government cannot justify the delay by arguing that it needed additional time after April 2013 or January 2015 to prepare a detailed report on religion or sort data about non-major religions. The delay in releasing religion data is particularly inexplicable because the time required to publish the data should decrease with the increasing use of information technology in census operations. Incidentally, the religion data from the March 1971 census were released in October 1972.

The delay reflects two deeper problems: growing political interference in the government’s statistical machinery and, possibly, a deepening communal crisis. This interference is reflected in the politically-motivated timing of the release of various datasets and reports as well as the disregard of expert advice on the design of data collection exercises, such as in the 2011 Socio-Economic and Caste Census.

Countries such as Lebanon, Pakistan and Nigeria have failed to conduct censuses regularly due to ethnic conflict and political instability. The hesitation to release data on religion may indicate a similar communal crisis is emerging in India. In some Indian states — such as Jammu and Kashmir, Assam, Nagaland and, until the 1980s, Punjab — the intensity of power struggles between communities has already affected the quality of government statistics.

The Hindu–Muslim statistical conflict in India began with the politicisation of religion and census during the colonial period. Muslims have feared the Hindu majority since the late 19th century; the Hindu fear of the fecund Muslim goes back to the first decade of the 20th century. The introduction of communal electorates in 1909 and the communal partition of Bengal in 1905 accentuated the political significance of demographic statistics. A number of influential pamphlets were published in the early-1900s, which continue to inspire propaganda in independent India. Interestingly, the same colonial census data that convinced Muslims that they were at the mercy of an unassailable Hindu majority, also convinced Hindus that they were a dwindling community soon to be eclipsed by Muslims.

In 1941, before the communal partition of British India, all communities inflated their headcounts in the undivided provinces of Punjab and Bengal. Eastern India also witnessed a struggle over the religious identity of tribes, who were seen as a swing community. After independence, Indian Punjab witnessed a protracted Hindu–Sikh power struggle disguised as a Punjabi–Hindi language conflict. The Kashmir Valley continues to be unprepared to accept any headcount that affects the Muslim majority status of Jammu and Kashmir or weakens the electoral dominance of Kashmiri Muslims in the state legislative assembly. Assamese (Hindus) have similar concerns regarding (Muslim) Bengalis. The ongoing debate on extending affirmative action — a quota system for public jobs, public university placings and elected assemblies — to scheduled caste Christians and Muslims is likely to further politicise the census.

Unfortunately, information technology and advanced statistical tools cannot resolve problems that have roots in a divisive political culture. Strengthening the autonomy of government’s statistical wing, though essential, is insufficient to address the problem. Other government bodies that contribute to transparency in the public sector, such as the Information Commission, need to be strengthened. The participation of non-governmental stakeholders needs to be encouraged both during data collection and dissemination. Unless these steps are taken community level data in India will continue to be amenable to politicisation, which ultimately cuts to the credibility of the government data in general.

Vikas Kumar teaches economics at Azim Premji University, Bengaluru.

SHARE: