From a new gross domestic product (GDP) series to a revised Index of Industrial Production and inflation indices, alongside fresh interventions including payroll reporting based on EPFO data, statistical measurement tools to gauge economic development underwent some change over the past four years. In the process, though, regular data sets, such as the Annual Employment-Unemployment Survey, have gone missing.
Additionally, even as the new GDP series based on 2011-12 propelled the growth rates to much higher levels than the 2004-05 series, the continuous flip-flop over the back series has resulted in a statistical vacuum about the comparative GDP growth rates for pre-2011-12 years, making it impossible to compare the new growth data with the growth during the UPA years. Here’s taking a stock of the various data sets that went missing or were modified over the past four years, alongside the ones that were newly added:
Missing data sets
Annual Employment-Unemployment Survey:
Six rounds of the Survey, which is conducted by the Labour Bureau under the Ministry of Labour and Employment, have been done so far but only five have been released, with the last one released in September 2016 for the period 2015-16.
In March 2018, Labour and Employment Minister Santosh Kumar Gangwar said that the survey has been discontinued based on the recommendations of the Task Force on Employment.
Government officials said that there were issues regarding the outsourcing of the data entry work related to the sixth Annual Employment-Unemployment Survey that has resulted in delay in its release. The survey has been conducted but work is underway on data compilation, they said. “Earlier, only some portion used to be outsourced such as feeding of data. Now, the whole data set (feeding and compilation) has been outsourced. It’s huge set of data so it’s taking time. About 1.60 lakh records are being compiled,” one of the officials said. The 2016-17 Survey would be crucial for doing a comparative analysis of the post-demonetisation period following the release of the Periodic Labour Force Survey which has been conducted for 2017-18. “The 2016-17 data is crucial because the next data point that you are going to get is 2017-18 which is the Periodic Labour Force Survey but there I have no basis to compare. We wouldn’t have anything for the previous year. So, 2016-17 is critical,” International Growth Centre’s India Director and former Chief Statistician of India, Pronab Sen, said.
GDP back series:
In January 2015, the government shifted to a new base year of 2011-12 from the earlier 2004-05 for national accounts. The base year of national accounts was last revised in January 2010. In the new series, the Central Statistics Office (CSO) did away with GDP at factor cost, adopting the international practice of valuing industry-wise estimates as gross value added (GVA) at basic prices.With the shift to the new base year, growth rate for 2013-14 was estimated at 6.9 per cent, while the earlier estimate on the 2004-05 base was 4.7 per cent. Similarly, the GDP growth rate for 2012-13 was revised upwards to 5.1 per cent compared with 4.5 per cent estimated earlier.
However, there were various flip-flops about the release of back series of GDP rates for years preceding 2011-12 linked to the new base year of 2011-12. Initially, after the release of the new series in January 2015, government officials said that the back series will be released by the year-end. In 2016, the CSO had then planned to release the back-series along with the release of provisional estimate of GDP for 2015-16 on May 31, failing which it then planned to release it by mid-June. In late 2016, a committee having Niti Aayog officials, finance ministry officials and economists as its members met to discuss the findings of the GDP series for the years preceding 2011-12. One of the members who attended the meetings said that the findings were questioned for pre-2011-12 years and the series was never publicly released.
In 2017, the then Chief Statistician of India, T C A Anant, said that the back series was proving to be a “major statistical challenge” as the database uses from the corporate affairs ministry was unavailable for previous years. He had said that “it’s not going to happen that quickly.” In March 2017, officials had indicated that the back series may not be released.
Government officials, however, now say that the work on back series has not been completely put on hold. An official said that the government may splice the GDP series to distinguish between the before and after MCA phase. “We are trying to work out an intermediate way that it could be in harmony with what was there earlier and whether we have to use splicing to go towards the new series. Whether we have to do the splicing, that’s what we have to take a call on,” the official said. “Splicing of data sets could be done to distinguish between pre and post MCA period. So, 2004-05 to 2010-11 maybe we will do one set, then from 2010-11-2011-12 and then 2011-12 onwards we have another set. But we have to take a call depending on the completeness of the database because even the MCA database is evolving and shifting from a very vague and loose system to a very robust system. We have to look at the data, that is slightly taking time,” the official said.
Modified data sets:
Quarterly Employment Survey:
Initiated in the aftermath of the 2008 global financial crisis to track its impact on employment, it underwent a revamp in 2016 after it met with criticism of being not the adequate representative of formal employment in India.
The Survey was started in October-December 2008 with approximately 3,000 units, which later declined to below 2,000 units by October-December 2015. In January-March 2016, the sample size was expanded to more than 10,000 units across eight sectors including manufacturing, construction, trade, transport, education, health, hotels and restaurants and business process outsourcing, covering about 81 per cent of establishments with more than 10 employees. Seven reports of the revamped QES have been released so far, with the last one released for July-September 2017 in March this year.
The Labour Ministry has now decided to hold the release of the eight round (October-December) of the Survey even though the report is almost ready. Ministry officials said that since the EPFO enrollment data is being regularly released, there is no need to release the Survey that anyway has statistical shortcomings.
The Task Force on Employment had pointed out several flaws in the QES regarding the sample size and frame of the survey. It said that since the majority of enterprises in India are small (less than 10 workers), in reality the sample of QES only represents about 1.37 per cent of all enterprises in the country or 21.15 per cent of non-agricultural employment. Economy-wide, the QES covers only about 2.77 crore workers out of a total of 47 crore or more workers in total, it said. Also, sample frame is not updated and until December 2015, sample was not random and therefore, it is not suitable for generating employment estimates, it said.
A panel headed by former Chief Statistician TCA Anant is looking into the merits of QES and whether it should be continued, a government official said. The panel is also looking at the frame needed for the survey since the frame is old as it is based on Sixth Economic Census (2013-14). A survey was also being planned for less than 10 workers incorporating information from households, though the technical group led by Anant is yet to take a final call on it. The survey would have been a part of the QES for less than 10 workers to capture data for informal sector based on information from households, the official said.
For the proposed survey of less than 10 workers, several factors are yet to be decided including sample size, manpower, funds, resources. It is also constrained by the limitation of how it can be clubbed with the survey for more than 10 as the sample frame of that survey is not updated until a new Economic Census is conducted.
Administrative data and data from government schemes:
The government has started releasing enrolment data for Employees Provident Fund Organisation (EPFO) as an indicator for formalisation of jobs in the country. As per the data, total net payroll enrolment stood at 39.35 lakh during September 2017-March 2018, the latest data released by EPFO showed. Payroll enrolment data was first released in April this year.
Economists have flagged several issues with the EPFO data being seen as a data for measuring job creation. The EPFO itself in the footnote of the data release said that the data for most recent months are provisional as updation of employees records is a continuous process and are likely to be updated in subsequent months. Also, though the members’ data above are linked to Aadhaar, the estimates may include temporary employees whose contributions may not be continuous for the entire year, the EPFO had said. In January 2018, Soumya Kanti Ghosh & Pulok Ghosh had published a study titled, “Towards a Payroll Reporting in India”, and estimated that around 70 lakh jobs might be generated in 2017-18.
With an aim to capture the impact of government schemes on employment, the government has already kick-started a survey on Pradhan Mantri MUDRA Yojana. “Work is already underway for that in the field and the report will come by 6 months or so,” the official said.
The Task Force on Employment in its report had said that, in principle, data from several government schemes and programmes such as the Mahatma Gandhi National Rural Employment Guarantee Act, Pradhan Mantri Grameen Sadak Yojana, Micro Units Development and Refinance Agency, Integrated Child Development Services programme, Housing for All, Pradhan Mantri Kaushal Vikas Yojana, Deen Dayal Upadhyaya Grameen Kaushalya Yojana, and various infrastructure projects and livelihood schemes create jobs. However, the major limitation in using data from these schemes is that, in some cases, additions to payrolls may not represent new jobs and simply indicate employment shifts, it had said.
Periodic Labour Force Survey (PLFS):
A new survey being conducted by NSSO. PLFS is a continuous/regular survey for generating estimates of various labour force indicators on quarterly basis for urban areas and annual basis for both rural and urban areas, at State/UT and all-India level. The fieldwork for this survey commenced on April 1, 2017 and is underway. The report is expected to come out by September-October this year, a government official said.