Wave 22 delivers more data in spite of lockdowns
by Mahesh Vyas
The twenty-second Wave of CMIE’s Consumer Pyramids Household Survey was completed on schedule on April 30, 2020. This Wave was hampered once again by restrictions on movement and the renewed need to maintain physical distances because of the second wave of the Covid-19 virus. Nevertheless, CPHS saw an increase in its sample size, an increase in the responding households and an increase in the response rate during Wave 22 compared to the past three Waves that were also impacted by the mobility restrictions.
The sample size was raised by 2,256 to 176,661 households. The CPHS sample has increased after having remained unchanged at 174,405 households for six consecutive Waves since Wave 16 of January-April 2019. Additions to the sample in Wave 22 were entirely in urban areas of Karnataka, Telangana and Andhra Pradesh. 832 households were added in Karnataka, 576 in Telangana and 848 in Andhra Pradesh. All these additions were to offset an increase in non-responses in the urban regions of these states because of a variety of reasons. The non-responding households have not been removed from the sample. Efforts to get responses from these households continue. There were no deletions from the sample in Wave 22.
Response to CPHS from the sample households were seriously affected when the government introduced Covid-19 related restrictions. Total responses that averaged over 147 thousand per Wave during 2019 fell to 112 thousand in the January-April 2020 Wave when the months of March and April 2020 came under lockdown restrictions. Usually CPHS generates 36-37 thousand responses per month. But, in March 2020, responses were down to 25,860. This was the first month when CPHS responses fell below the 30,000 mark in a month. They fell further to their lowest level of 13,294 in April 2020 before embarking on a slow journey of recovery. By March 2021, responses had covered substantial ground. CPHS accepted 34,681 valid responses in March. But, local lockdowns and other restrictions brought the responses down to 26,952 in April 2021.
CPdx users stand to benefit from a healthy increase in responses during Wave 22. They rose by 6,553 households’ from 123,188 responses in Wave 21 of September-December 2020 to 129,741 in Wave 22 of January - April 2021. Urban responses increased by 3,523 from 84,831 to 88,354 and rural responses increased by 3,030 households from 38,357 to 41,387.
The increase in responses in Wave 22 was more than the increase in the sample during the Wave. As a result, the response rate increased from 70.6 per cent in Wave 21 to 73.4 per cent in Wave 22.
The response rate improved in both, rural and urban regions. But, the response rate in rural India continues to remain much lower than the response in urban regions. The response rate in urban India improved from 76.4 per cent in Wave 21 to 78 per cent in Wave 22. In rural India the improvement was much bigger from 60.5 per cent in Wave 21 to 65.3 per cent in Wave 22. But, the response rate is much lower in rural regions.
Responses have been poor in Sikkim (27.7 per cent), Jammu & Kashmir (29 per cent), Odisha (42 per cent), Goa (46.5 per cent) and West Bengal (46.8 per cent). Other below-average states were Tamil Nadu (59.8 per cent), Assam (65 per cent), Telangana (66.1 per cent), Karnataka (67 per cent), Chhattisgarh (68 per cent), Andhra Pradesh (69 per cent), Gujarat (70.7 per cent) and Maharashtra (72.4 per cent).
Mobility restrictions imposed after the Covid-19 pandemic led to CMIE shifting its survey operations from face-to-face interviews to telephonic interviews. CPHS is a face-to-face survey conducted by interviewers visiting the sample households and interviewing members of sample households in person. Restrictions on movement and the need to maintain physical distancing following the outbreak of the Covid-19 virus have limited these visits. During Wave 19 (January-April 2020), six of the sixteen weeks were impacted by the lockdown. Responses fell during this Wave and, of the accepted records, 13 per cent were based on telephonic interviews. The proportion of telephonic interviews increased to 45 per cent in Wave 20 (May August 2020). During Wave 21, only 6.3 per cent of the accepted records were based on telephonic interviews. And, in Wave 22, telephonic interviews accounted for only 2.4 per cent of all accepted records.
Mobility restrictions were the pre-dominant reason for low responses. Because of these restrictions the count of households that could not be executed increased sharply during the period of mobility restrictions. These restrictions are not limited to a complete nation-wide or a complete local lockdown. There are restrictions even when there are no lockdowns. These include lack of availability of public transport and lack of availability of other services associated with travel such as restaurants, boarding and lodging facilities, etc. even when there are no complete lockdowns. Face-to-face interviews were also curtailed because of the need to maintain physical distance.
Inability to execute the survey accounted for about 4,600 non-responses from 2014 through 2016. These increased to over 6,000 in 2017 and then to about 10,000 in 2018 and 2019. The share of non-response because of non-execution of the survey rose from about 2 per cent initially to about 6 per cent later.
After the imposition of the lockdown in late March 2020, the share of non-response due to non-execution has risen substantially. In Wave 19 (January-April 2020) when the lockdown impacted execution partly in March and entirely in April, this share rose to 24 per cent. In Wave 20 that bore the biggest brunt of the lockdown, non-response due to non-execution shot up to 42 per cent. It then came down to 21.5 per cent in Wave 21 as restrictions eased further.
Mid-way through Wave 22, India experienced the second wave of Covid-19 infections. During January and February 2021, non-response due to non-execution was down to 16 per cent and 14 per cent, respectively. It inched up a bit to 15 per cent in March and then shot up to over 32 per cent in April 2021. As a result, Wave 22 ended with a non-response due to non-execution of 19.6 per cent of the total sample. Note that this is lower than the 21.5 per cent pencilled in Wave 21. But, were it not for the second wave of Covid-19, the record could have been much better.
Restrictions because of the second wave of Covid-19 have been milder than in the first wave in 2020. Partly because of this and partly because of the preparedness, CPHS ended Wave 22 with more observations than in any wave since the pandemic first hit India.