List of Contents
Source: The post is based on the article “Is there a rural bias in national surveys?” published in The Hindu on 25th July 2023.
Syllabus: GS 2 – Government Policies and Interventions
Relevance: concerns associated with national surveys
News: A panel headed by Pronab Sen, a former chief statistician of India, was recently formed by the Indian government to review the National Statistical Organization’s (NSO) methodology.
Why did the government form the panel?
The government has formed a panel after the shortcomings highlighted by Shamika Ravi and Bibek Debroy in their articles.
They argue that outdated survey methodology used by the National Sample Survey (NSS), National Family Health Survey (NFHS), and Periodic Labour Force Survey (PLFS) have consistently underestimated India’s progress.
They believe the Indian economy has been dynamic in the last 30 years, and the current methodology fails to capture this reality.
Why is it important to review the methodology used by the National Statistical Organization (NSO)?
National level data are essential for research, policies, and development. Therefore, it is crucial to thoroughly review and assess the assertions made by various specialists regarding the accuracy of this data.
Moreover, in order to analyze the claims, the article tries to focus on NFHS data.
Does the NFHS have a rural bias?
Ms. Ravi and Mr. Debroy have argued that NFHS which depend heavily on the last Census data, systematically overestimates the rural population.
However, an examination of five rounds of NFHS data contradicts this claim, showing no systematic bias towards the rural population.
When the NFHS estimates of urban population are compared to World Bank estimates and urban percentage projections from Census statistics from 1991, 2001, and 2011, no indication of systemic rural bias can be found.
On the other hand, it has been found that NFHS-3 underestimated the rural population, while NFHS-2 and NFHS-5 may have overestimated it. However, these errors appear to be random rather than systematic.
How can these errors be minimized?
Urban areas tend to have higher percentages of no-response compared to rural areas. However, this is not related to either rural or urban bias in estimation.
An analysis of the percentages of the urban sample in the unweighted sample suggests that giving proper weights may assist in significantly addressing the errors.
Hence, by appropriately assigning sample weights after considering all potential sources of error, the underrepresentation of rural or urban areas can be significantly rectified.
What can be the way ahead?
The Pronab Sen Committee should prioritize addressing concerns regarding sample representation rather than completely overhauling the survey methodology. Otherwise, there is a risk of introducing a systematic urban bias in policymaking, planning, and financing.