India has historically had a powerful statistical system. However, the global pandemic that pushed India into a nationwide lockdown has disrupted not only normal life but also the collection of statistics.
It is important to understand the shift the consecutive lockdowns in multiple states have created in the macroeconomic data and whether the statistical system is robust to face the unprecedented challenges created by unexpected events such as lockdowns and curfews.
To discuss the multiple themes and subtopics to assess the impact on Indian macroeconomic statistical systems GenAlpha Data Center at IMPRI Impact and Policy Research Institute, New Delhi, hosted #WebPolicyTalk-The State of Statistics #DataDiscourses, special lecture, Evaluating Indian Macroeconomic Statistical Systems: Data Trends, Challenges for Evidence and Impact amid COVID-19 on 11 May, 2021, with Dr J Dennis Rajakumar, Director of Economic and Political Weekly Research Foundation, as the guest speaker.
Dr Arjun Kumar, Director IMPRI, opened the session by introducing the guest speakers and the discussants. The discussion was handed over to Dr Rajakumar for his opening remarks.
Dr Rajakumar started his talk by raising the difficulty of covering a wide range of topics. It is difficult to define a criterion to evaluate the Indian Macroeconomic Statistical System. Therefore, he based his talk on four major macroeconomic indicators: GDP, IIP, CPI and WPI.
He would be evaluating the impact of the lockdown and COVID-19 pandemic on these four indicators based on press releases. According to him, a robust statistical system must ensure continuity and an uninterrupted system for collecting statistics. Therefore, he wants to evaluate during his presentation how the statistical system took cognisance during the past year.
With nationwide lockdown being imposed, it disrupted the process of collecting data, especially indices such as CPI and WPI, which require data to be collected in primary format. It also affected the field visits, which ensured a regular flow of information.
Due to the extension provided by the regulatory bodies to submit information on financial returns, the usual data sources were substituted by GST or interacting with professional bodies. These methods, however, proved to be limited in nature.
The period of the pandemic has proven to affect the quality, flow and availability of information and CSO (Central Statistical Organisation) faced challenges with specific sectors or indices.
While talking about GDP-related releases, Dr Rajakumar explained the kind of estimates that were released like quarterly estimates, advanced estimated and revised estimates. Quarterly and advanced estimates were affected during the pandemic. The first two quarters showed a negative GDP growth rate.
Looking closely at annual estimates, in the first advanced estimates produced GDP at market price was expected to fall -7.7% and the second advanced estimated predicted to fall to -8.0%. This is attributed to the heavy subsidies provided by the government to certain sectors and the reliefs provided to the banking system, which makes the GDP at market prices differ drastically from GVA at Basic Prices.
Even with a lockdown, the release date of the Index of Industrial Production was never missed. However, the lockdown did affect the response rate. Since April 2020, the response rates have been improving and the first revision (T+1) estimates also have been improving up to January 2021. The final revision (T+2) has also improved over the months.
However, looking closely at the IIP General Index except for the months of March, April and May, other months have not seen much change.
Dr Rajakumar had taken the average of IIP General Index for the quarters and noticed that the difference between provisional estimates and final revisions was minor. While analysing the monthly data, the IIP values were more prominent in the pandemic’s initial months, and the NSSO found it appropriate not to calculate the growth rate.
In one of the press releases, it was commented, “It is not appropriate to compare the IIP of April 2020 with earlier months and users may like to observe the changes in IIP in the following months.” This was a deliberate step as most industries were not operating from the end of March 2020.
The NSO also releases the Consumer Price Index (CPI). CPI is collected by 1,114 urban markets and 1,181 villages by personal visits by field staff on a weekly roster. On 19 March, the personal visits of these field staff were stopped. For the month of April, price data was collected through telephonic inquiries from designated outlets along with the personal purchase visits done by field staff.
The state-level CPI was not prepared for the month of April and May. In May, the NSSO adopted the Business Continuity Guidelines for imputing the missing sub-group index with an index computed using the index at a higher level of aggregation.
The Wholesale Price Index (WPI) is released in two formats, provisional and final. The price data is collected from various institutional sources and industrial establishments spread across the country through the web-based portal National Informatics Centre (NIC). The impact on WPI caused the Price collection of Manufactured Products through personal visits of price collectors to be suspended with effect from 19 March 2020.
The price movement sub-groups were worked out by taking the prices of only those items for which at least 25% of price quotations were reported from selected manufacturing units.
Dr Rajakumar had presented a few concluding remarks on the analysis of the different indicators mentioned before.
“The statistical system has remained resilient in times of crisis.”
Even though the process of collecting data was interrupted, the timeliness of the release of the reports was intact and they have been quick to adapt to an internationally adopted methodology for estimating missing values.
The crisis has also caused the statistical system to be more transparent with the reporting of response rate and percentage of markets covered. Dr Rajakumar believes that the present system can help improve the competency of the system. He also suggested a new outlook at determining CPI by using data from the online sales of goods or services, making it more reliable.
Dr Rajeswari Sengupta, Assistant Professor, Economics, Indira Gandhi Institute of Development & Research (IGIDR), Mumbai, focused on trust and confidence in the official statistics system.
She stated that after independence, the Indian statistical system received widespread appreciation. however, in the last 10 years, there has been severe erosion of trust and confidence in the official statistical system
The Indian economy has experienced several episodes of fluctuations and volatilities in economies in the last 10 years.
Several corruption scandals from governments, the protracted crisis in banking and private corporate sectors, demonetisation, and patchy implementation of goods and services tax have resulted in a shrinking economy.
She further highlighted that due to all these growing volatilities, the development aggravated the need and urgency among all kinds of stakeholders the demand for data to understand how the economy and its associate sectors are performing.
In the middle of the pandemic, when people are more desperate to understand the state of the economy and what kind of recovery it’s going to experience, there is growing reliance among stakeholders on unofficial data and high-frequency data. Hence, a lot of data coming today is from private organisations.
Today there exist parallel data sources coming from national and private systems resulting in divergence of statistical data and the official statistical system has no longer monopoly on data.
There is a need for an honest and holistic approach to resolving this massive crisis of trust and confidence in the National Statistical System.
She further commented that several steps were needed to build reliability and credibility in official data. First, much greater transparency is required in methods used for calculating statistical data. Also, there should be regular engagement with stakeholders and critics.
It is very important to revive the National Statistical Commission as an independent body.
Talking on MIS (Management Information System), Dr Dennis commented that it does not happen to be a system that the planning commission had. Thus, just referring to the data coming from the dashboard is not feasible. Also, MCA data set 2020 is just a financial data set submitted by companies. Thus, it creates serious doubts whether we can use these data sets or not.
Commenting on building up SDP, he said that today we do not follow the bottom-up approach; rather, we follow a top-down approach which means there are estimates for the country as a whole which is then apportioned. The system of apportioning accounts for roughly around 70-75% of SDP.
Sharing views on quarterly estimates, he stated that there was a need to look at estimates from the point of view of not just understanding directions. If the rates and ratios are generated at a more frequency level, it will lead to a more robust data system. He further emphasised the need for the absolute size of GDP estimated even at the sectoral level.
Dr P C Mohanan, Former Chairman, National Statistical Commission, MoSPI, highlighted the importance of data in crisis situations.
He commented that there exists unavailability of appropriate data while for having consistency, the statistical system follows standardised procedures. On the brighter side, transparency has improved in terms of reporting.
We need a flexible statistical system where we can think of some innovative methods of obtaining data.
Acknowledgement: Arjun Sujit Varma is a Research Intern at IMPRI