Subodh Varma, Sep 27 2014 http://timesofindia.indiatimes.com/india/28-million-people-in-Census-blind-spot/articleshow/43579291.cms NEW DELHI: The Census of India in 2011 left out about 28 million persons from its headcount. This is the startling conclusion of a survey carried out by the Census people themselves, a few months after the enumeration work finished in March 2011. Called the Post Enumeration Survey (PES), it was a near repeat of the Census, except that it was carried out to cover only about 4 lakh persons across the country.
Big data is a strange beast. But the strangest thing about it is that it's all relative - what you think is big is smaller than what is bigger, and vice versa. The 2011 census counted over 1.2 billion Indians, collecting about 30 bits of individual information from them, besides 35 other queries about their households. That's big data. Compared to this, the 28-million uncounted Indians make up just 2.3 percent of the total population. This small share means that there is no cause for alarm, the Census office says. Nothing will have to be changed. In 2001, the net omission rate was roughly the same, although it was lower in 1991 and 1981. How did so many people go uncounted in a rigorously designed survey like the Census? "Most of this occurs because of hard-to-reach people, particularly people with no fixed residence such as street people, and because of people who are travelling," explains Pronob Sen, chairperson of the National Statistical Commission. "An estimated coverage omission rate of 2.3% is by no means unusual or excessive even if it adds up to 2.8 crore people. It does not really lead to any serious bias. If the omission rate was higher than 5%, then a more substantive survey would have to be done to correct the Census estimates. To put it in perspective, the NSSO surveys underestimate population by a much higher proportion (6 to 12%)," he told TOI. Among the uncounted, some strange facts emerged. The share of people left out in the Census was higher in some states. In the central zone made up of UP, Uttarakhand, MP, Chhattisgarh, as many as 4.2 percent of people were omitted. That's close to the statistically permissible limit of 5 percent that Sen mentions. Another strange fact is that omission rates are generally higher in urban areas as compared to rural areas. But the most bizarre finding is this: fathers and mothers of the head of the household were more likely to be omitted from the Census count, just as were other relatives. But the worst were unrelated persons staying with the family, like servants. Over 15 percent of these were not counted. Besides these variations, the PES tested out the variation in other data like educational level, employment, disability, marital status etc. on a smaller subset of 20 percent of the original sample. This is where the good news came from: there was virtually negligible difference between Census and PES results. So, the Census figures will go down in history as reported, forgetting the 2.8 million forever. -- Avinash Shahi Doctoral student at Centre for Law and Governance JNU Register at the dedicated AccessIndia list for discussing accessibility of mobile phones / Tabs on: http://mail.accessindia.org.in/mailman/listinfo/mobile.accessindia_accessindia.org.in Search for old postings at: http://www.mail-archive.com/[email protected]/ To unsubscribe send a message to [email protected] with the subject unsubscribe. To change your subscription to digest mode or make any other changes, please visit the list home page at http://accessindia.org.in/mailman/listinfo/accessindia_accessindia.org.in Disclaimer: 1. Contents of the mails, factual, or otherwise, reflect the thinking of the person sending the mail and AI in no way relates itself to its veracity; 2. AI cannot be held liable for any commission/omission based on the mails sent through this mailing list..
