Opinion
The spatial spread of economic activities must be mapped for data accuracy
Summary
- Governments at the state level must first acknowledge the crisis in statistical systems and then chalk out reforms to rebuild them.
How well an average person does in life is determined largely by an accident of birth. A child born in rural Tamil Nadu is better nourished and educated, on average, than one born in rural Chhattisgarh. There are stark differences in economic outcomes even within states. Economic opportunities are far more abundant in the Konkan belt of Maharashtra than in Marathwada.
Yet, we do a poor job of tracking regional inequality. When the national accounts estimates are released each quarter, one has no way of knowing how much each state or region has contributed to the national total. The few states that do compile such estimates do not release them publicly. Even annual estimates of state domestic product (SDP) are issued after a long time lag. District domestic product (DDP) estimates depend on guesswork to a large extent, and aren’t comparable across states.
The neglect of local statistical capacity in India has a long history. As India’s modern statistical edifice was being erected by the British in the mid-19th century, the Bombay government requested funds to set up a provincial statistical office. The central authorities denied that request, retarding local statistical capacity. The lack of investment in local statistical offices meant that by the middle of the 20th century, only three provinces (Bombay, Uttar Pradesh and Bihar) had any estimates of provincial income.
It was only after India’s independence that estimating state income became a priority. Following recommendations of the National Income Committee in the early 1950s, a handful of states began preparing state income estimates. India’s second five-year plan funded the establishment of statistical bureaus in other states, which began generating state income estimates. Since the methods and sources differed across states, the Planning Commission and Finance Commission demanded that the Central Statistical Organization produce comparable SDP estimates. This was meant to be a short-term measure till the state estimates were improved and standardized. In the 1970s, a committee led by national accounting pioneer Moni Mukherjee prepared a template for standardized regional accounts. Some recommendations were acted upon, several others ignored. Each Finance Commission still relies on “comparable" SDP estimates produced by the Union ministry of statistics and programme implementation (Mospi).
In most states, the directorates of economics and statistics (DES) are headed by uninterested civil servants who may have preferred other jobs or by statisticians who do not rank high in the state administrative hierarchy. Lack of attention and funding from state governments has made these DES highly dependent on Mospi. Lately, even Mospi’s support has been found wanting.
Several flagship surveys of Mospi are now run without involving state DES. Even the latest economic census—traditionally run by DES officials—involved a marginal role for the DES. Mospi deployed staff from Common Service Centres (CSC) run by the IT ministry to collect data. Some DES officials opposed the move, arguing that CSC staff lacked exposure to basic statistics. The results from the field seem to have vindicated their stance. Seventeen of 28 state-level coordination committees are yet to approve the census findings, and a majority of them have raised concerns over data quality, Mospi said in response to an RTI query from this writer.
SDP estimation has also been a bone of contention since the 2015 base-year revision. The new series relied heavily on indirect methods of estimating SDP figures from national totals. A 2017 research paper by Gujarat DES official Manish Pandya and economist Ravindra Dholakia argued that this led to inaccurate estimates of state incomes. Mospi found it hard to evade the questions being raised in the Prime Minister’s home state. It set up a panel to review SDP estimates, and asked Dholakia to chair it.
The Dholakia committee report in 2020 recommended that both state and national incomes should be estimated using a bottom-up approach in the next base-year revision. This would improve the accuracy of state as well as national totals. The Mukherjee committee had made similar arguments. Yet, such recommendations can be effective only if state statistical bureaus have enough heft to generate high-quality granular data. Without support from officials at Mospi and state secretariats, they would continue to struggle. DES staff also need help from outside experts to incorporate new tools (including geospatial ones) in their work.
Acknowledging a crisis in state statistical systems is the first step towards solving it. In 2020, the Madhya Pradesh (MP) government set up a statistical task force headed by the former National Statistical Commission member Amitabh Kundu to review its statistical system. The task force report is an eye-opening account of the deep rot in the state’s statistical system. It noted a lack of any effective mechanism at the state or district level to check the quality of data-sets compiled in different departments, and that some estimates were being calculated “manually" rather than by using computers.
As recommended, MP set up a state statistical commission to provide technical guidance to the DES and review key statistical activities. Other states need to follow its lead and chart a roadmap for revamping their statistical systems. Without it, don’t expect granular high-quality data-sets from states.
This is the concluding part of a three-part series on measuring economic inequality.
