Concerns over the sampling design of national surveys not tenable: Govt

Objections raised by various experts on the sampling design of the national surveys like the National Sample Survey (NSS), National Family Health Survey (NFHS), and Periodic Labour Force Survey (PLFS) are not tenable, as the conclusions drawn about the sampling are based on faulty approaches, Rao Inderjit Singh, minister of state, Ministry of Statistics and Programme Implementation (MoSPI) told Parliament on Wednesday.

“The concerns raised by experts over the representativeness of NSS were brought to the notice of the ministry and the same has been examined in consultation with domain experts and reputed statisticians. It has been found that the concerns raised on the NSS sampling design are not tenable, since the conclusions drawn have been based on the assessment of sampling design with faulty application inapplicable method/approach,” the minister told the Lok Sabha in response to a question on the inaccuracies in the sampling of the major national surveys.

Besides, the minister pointed out that the ministry follows guidelines specified in the UN Handbook titled, ‘Designing Household Survey Samples: Practical Guidelines’ to design survey instruments and decide on the sampling method to be adopted. Also, the ministry uses the Relative Standard Error (RSE) method to test the robustness of its surveys.

“The methodologies followed in various NSS surveys have been finalised under the guidance of committees of experts and are regularly updated depending on the objective of the survey, availability of resources and other operational aspects,” he added.

Earlier in July last year, Shamika Ravi, member, PM Economic Advisory Council (EAC-PM) wrote in an article that all major surveys in India that were conducted post-2011, and used the Census 2011 for the sampling frame, have overestimated the proportion of the rural population significantly.

“This is one of the several problems with data quality, but it is a critical concern — and appropriately highlights the problem at hand,” she added.

Chairperson of the revamped Standing Committee on Statistics (SCoS) Pronob Sen told Business Standard that though the SCoS was not involved in scrutinising the claims of Shamika Ravi, the problems that she highlighted were essentially a sampling problem.

“The surveys only follow the statutory definition while classifying rural and urban areas. Thus, although all surveys use the census as the sampling frame, census towns are treated as part of the rural sector and are included in the rural sample and this is what makes rural representation in surveys look higher,” he added.

In his parliamentary response, the minister also specified that the department was using the updated sampling frame from the Urban Frame Survey 2017-22 to select urban samples, while it relied on census reports for rural samples, which excludes villages added to urban samples.

“During surveys, within each of the selected final village or urban block, all the households are listed and this serves as the sampling frame for the selection of ultimate stage units with the latest population as of the date of survey,” he noted.

The minister also pointed out that the National Family Health Survey (NFHS) conducted by the Ministry of Health and Family Welfare had undergone revisions in each round to provide relevant data in consonance with Sustainable Development Goal indicators and over the different rounds, the sampling design has been modified to cater to the need for providing estimates at not only the national and state level but also the district level.