WordPress database error: [Table './geograph_gny/gnywordpress_usermeta' is marked as crashed and should be repaired]
SELECT user_id, meta_key, meta_value FROM gnywordpress_usermeta WHERE user_id IN (18) ORDER BY umeta_id ASC

census of india

Census of India – A Critique

By: Mahendra K Premi
The Indian Census has been providing important and useful information on various demographic, social and economic aspects of the Indian society. However, the paper outlines definitional and other changes and how they may have a limiting influence on data usage.

Modern Census-taking in India started in 1872, has an unbroken record of population count every ten years, and the 2011 Census of India is the fifteenth in the series. Census of India has been providing very important and useful information on various demographic, social and economic aspects of the Indian society along with its living conditions and is marked by several milestones since its inception. The 1961 Census introduced a houselist schedule; it changed the concept of ‘economic activity’ from ‘income’ to ‘work’ approach and it decided to collect migration data down to each administrative unit rather than limiting it to district level as was done until the 1951 Census. These changes made a significant difference in the understanding of Indian realities. Since then every Census of India has introduced new questions and new formats.

The 2011 Census of India too has also been unique in several ways. It canvassed a new schedule termed National Population Register (NPR) to collect data on every citizen in India to introduce unique identity numbers (UID). Instead of asking for ‘age at last birthday,’ data has been collected on exact ‘date of birth’ both in NPR and in the population Census. Interestingly, the question on ‘nationality’ was specifically canvassed in NPR, while it was not included in the population Census questionnaire after 1961. The earlier ‘houselist schedule’ has been expanded and termed ‘house listing and housing census schedule’. While it has almost all the questions of the 2001 Census, it has additional questions on possession of mobile phone as also on computer/laptop with/without internet facility. Data tabulated from the above schedule provides valuable information down to the district/city level about the quality of housing on one hand, and socio-economic status of the household on the other.

Another unique aspect of 2011 Census of India is that the total population count data as well as several questions on age, literacy, work status etc. have been coded into three categories male (1); female (2), and ‘others’ (3) (for persons who report themselves as transgender); instead of (1) and (2) as in earlier Censuses.

The immediate role of the Indian Census in context of the Indian federal polity has been to determine state wise distribution of seats in Parliament resulting in increasing share of Uttar Pradesh, Bihar, Haryana, Madhya Pradesh, etc. and certain states like Kerala and Tamil Nadu losing their representation. Therefore, in 1976 at the time of presenting ‘Draft National Population Policy’, state wise seat distribution was frozen at the 1971 level for 25 years. In 2001 it was again frozen for the next 25 years to maintain a status quo of seat distribution. Census of India data has also been used to demarcate boundaries of Parliament constituencies and deciding whether a particular constituency is in general category or is reserved for a person belonging to the scheduled caste (SC)/scheduled tribe (ST). This exercise, based on 2001 Census, was done recently for the Lok Sabha, state legislatures – down to the Zila Parishads.

Population data helps planners understand how forces such as population growth, urbanisation and migration, work force structure and ageing etc. affect people now and in future. With the availability of  ‘age structure’, Census of India data have been used for making population projections for the nation and states, which have been used in national planning and policy making. These data have been further utilised in computing per capita national income and employment rates. Population Census data have regularly been supplied to UN Statistical Office for global population estimates and comparisons. They are also used by different international agencies for making comparisons on development indicators such as literacy, economic activity etc.

Census of India data reveal important information on region wise population distribution and economy. From the latter viewpoint, at the first level of disaggregation, the economy is divided into primary, secondary and tertiary sectors. As employment pattern of rural and urban workers is very different, economists analyse the changing pattern of work participation rates. They examine the classification of workers according to the nature of industry and their occupational category besides other detailed analysis.

Administrators, social scientists and researchers use Census data to study demographic changes taking place over time in respect to population size and growth, rural-urban composition, sex and age structure, migration, etc. Sociologists use the data to understand changes in social systems based on tribes, castes, religious/ethnic groups, etc. that are used to differentiate them from one another. Besides providing information on religious distribution by marital status, literacy and work participation, Census data aids geographers in understanding changes in population distribution at the state and district levels. However, few political scientists have utilised Census of India data for the understanding of, for example, impact of lowering voting age from 21 to 18 years; or analysing the effects of revision of constituency boundaries and their reclassification into reserved or general category.

However, it should be borne in mind that the Census count does not come without its limitations despite the above benefits. For instance, in the 1991 Census, there was substantial undercounting in Bihar resulting in a low decadal growth rate of 23.4 per cent during 1981-91. This, however, became 28.4 per cent during 1991-2001 simply due to a better count in the latter Census. Again, the decadal growth rate of Jains, changed from 23.7 in 1971-81 to mere 4.1 in 1981-91 and then to 25.9 per cent in 1991-2001 because of special attention paid to the question on religion by the then Registrar General, Shri Banthia.

Census of India data should be analysed and interpreted with caution. The question on economic activity has not been exactly the same from one Census to the next. For example, female workforce participation rate in Punjab was a mere 5.3 in the 1991 Census, the lowest in the country but that jumped to 21 per cent in the 2001 one due to the Technical Advisory Committee for 2001 Census emphasising animal husbandry while collecting data on economic activity followed by the enumerators considering a woman as working even if the household had one cow or buffalo. Later in 2003 it was established through various studies that almost every household has cattle, in rural Punjab. Therefore, time series Census of India data should be closely examined for definitional changes from one Census of India to the next.



Census of India data collection is a humongous task for Census enumerators. Despite the excellent training they tend to assume answers and skip questions and many anomalies still remain in the data collection methods as well as the type of information collected and the level at which it is collected. Also, despite the Census collecting a mine of information on the different aspects of population, hardly 20 per cent of it is published and utilised for analysis. As a word of caution, it is important to know that Census of India data for very small geographical units like village, or town/city and their wards, occasionally have errors which arise either at the enumeration, data analysis or at the printing stage. Many of those errors cancel out, or reduce in intensity, when the data are aggregated for large units. Consequently, it would always be useful to use Census of India data for districts or still bigger geographical units. If, however, one must use micro-level data, it should be used with great caution.

Leave a Reply

Your email address will not be published. Required fields are marked *