Big data from Government schemes, though painstakingly collated, is often useless, says AMEE MISRA
An important plank of Prime Minister Narendra Modi’s economic vision is decentralisation of power and greater empowernment of the States. However, the success of decentralisation hinges on the ability of the lowest levels of Government to quantify and manage their needs and capabilities.
Governments need timely and reliable facts to ensure correct resource allocation and design relevant policies. In the absence of efficient and effective data management systems in India, it is no surprise that countless state schemes in India fail to achieve their objectives. Given the need to converge massive bits of information into comparable, numeric formats, the importance of data management in Government can’t be overemphasised.
A recent Government of India document looking to evaluate the annual work plan of Chhattisgarh Government’s mid-day meal scheme notes that there are wide discrepancies in scheme-related numbers captured by the State’s management information system and those reported by the State Government in its annual work plan.
It notes that even though nearly 100 per cent of data required under the MDM-MIS portal has been compiled, there were wide variations particularly in the coverage of children, engagement of cook-cum-helpers, payment of honorarium to cook-cum-helpers, utilisation of cooking cost, utilisation of foodgrains, availability of kitchen-cum-stores, availability of drinking water and toilet facilities.
The Sarva Siksha Abhiyan website proclaims that “A computerised annual educational MIS system (DISE) is operational in the country.” In theory, the idea behind putting in place a sophisticated MIS is that school-level data can be collected, computerised and collated upwards to the Block, District and eventually State level — ensuring that State-level plans are an aggregation of school level needs. In reality this is rarely so and the Chhattisgarh case is hardly an exception.
It is mandatory that data be entered at all levels, be compiled and sent upwards in a timely manner in pre-determined formats. Both time and money are intensively engaged at all levels of Government to ensure that this is indeed so. No stone is left unturned in making sure we get the most diligently filled excel sheets where the numbers all add up.
But to what endIJ The data is seldom used for or even relevant to Government planning which is often a separate exercise. At best, the data finds its way in to published reports, where it is referred to for academic papers, published with disclaimers.
So, why is inconsistency and often uselessness of data such a fundamental feature running across state programmes and schemesIJ The answers lie in the nature of Government functioning. The capture, analysis, sharing, storage, transfer and visualization of programme data are all viewed as exercises in isolation. Setting up the database is seen as an end in itself, with nearly no emphasis on its use and integration into the larger planning process.
The eventual planning exercise carried out at the State or even District level makes little or no reference to numbers collated from the school level upwards. Besides, where a State Government will open another elementary school or appoint another teacher is often not so much a function of the relative needs of the region or the school, but that of political compulsions. No wonder schools and districts rarely take the process seriously.
As the deadline for data submission approaches, school, block and district officers will often shut themselves in a room with a bunch of last year’s excel sheets, updating them based on their “experience and understanding” and hopefully informed guess-estimates.
Given the size of the budget available for this exercise, the process may be fuelled by an endless stream of chai and snacks, perhaps even carried out over a period of few days on a retreat, in hotel rooms of State capitals where all officers will gather. The casual approach comes directly from the fact that those engaged in filling the numbers are fully aware that no use will be made of these numbers going forward. This sets in place a vicious circle of irrelevant data.
A 2013 report by the erstwhile Planning Commission’s think tank found that Government data across most schemes was either incomplete or inconsistent. It observes that an effective MIS should take governance away from an overemphasis on “rule based approaches that primarily focused on process regulation, in compliance with centrally prescribed standards and rules”.
However, the way the MIS is designed and executed in India it does precisely this. Even when an officer is evaluated for his work for filling up the data sheets, what is assessed is whether a completed form or data sheet was sent in time and not whether this data was of any use or even accurate to begin with. There are no cross checks and no analysis of data coming in over a period of time to evaluate performance. To put it simply in the planning jargon – while the ‘output’ has been taken care of, the eventual ‘outcome’ is ignored.
This isn’t merely the case with education or other scheme-related data. large and often embarrassing inconsistencies have often been found in Government macro-economic data too, such as GDP growth rates, wholesale price indices and Index of Industrial Production numbers. While there exist simpler and direct explanations for all these inconsistencies (double counting, changes in categorisation from one year to another etc), the fundamental nature of the Government data collection and use process underlying all of these remain. Everyone does their task and moves on, with little or no integration into the larger picture.
(The writer is senior research analyst at CPR's Accountability Initiative which seeks to improve the quality of India’s public services by promoting informed and accountable governance)