India’s data-driven governance regime is not foolproof
It serves as a reminder of the dangers and vulnerabilities of data when they are deployed without adequate investments in quality, objectivity and credibility. This is now a challenge for the new government.Updated: Mar 27, 2019 18:20 IST
The ongoing controversy over the credibility and objectivity of India’s statistical system is one of the most damaging legacies of this out-going government. But here’s the irony. Even as India’s economic statistical machinery has been brought to the brink of crisis, in recent years, an effort has been underway to develop a new data-driven governance regime, especially for social sectors. This new regime relies largely on administrative data, which is routinely collected by government departments to track and monitor the progress of schemes and is increasingly beginning to shape day-to-day administration. This will have significant consequences both on the dynamics of scheme implementation at the grassroots as well as public understanding of what governments achieve. Given its import, this new data regime requires as much careful scrutiny and public debate for what it reports (and suppresses) as India’s economic statistics have recently generated.
This penchant for collecting administrative data to monitor scheme progress, at scale, predates this government. Arguably, the first attempt to create a 21st century technology-based monitoring system or Management Information System (MIS) was made with the launch of the National Rural Employment Guarantee Act in 2006. But the large-scale adoption of technology tools by this government has accelerated these efforts. Economists Abhijit Banerjee and Shrayana Bhattacharya in a recent article in The Indian Express note the existence of as many as 400 MIS’s for monitoring the Direct Benefit Transfer (DBT) schemes, a number far beyond what most middle income countries report. Technology has also enabled administrative data to become increasingly more sophisticated. GIS maps, data dashboards and mobile apps are today the new shiny tools for collecting and tracking progress.
But beyond the creation of new data systems, what is unique about this government, compared with its predecessors, is the attempt to deploy data as a tool for centralised monitoring and administrative action.. The Niti Aayog has taken the lead in using this data to develop indices that rank states on health, education, water and sustainable development-related indicators. The Aayog’s flagship scheme — the Aspirational Districts Programme — is now using administrative data to rank a cohort of backward districts on progress made on key social indicators.
By all accounts, Prime Minister Narendra Modi tracks progress in these districts using these indicators, resulting in a frantic race to the top. Not to be left behind, various line departments in New Delhi have created their own rankings — from the Swachchata Sarveskhan to the ease of living index. These rankings are often cited as evidence of India’s progress toward the goal of “competitive” federalism. Going forward, data and rankings of this nature may well impact the financial health of state governments. The terms of reference of the Fifteenth Finance Commission mandate the commission to consider developing a performance-based financing structure that links fiscal transfers to performance against flagship schemes and progress on Sustainable Development Goals.
On first principles alone, this new data regime is an important step in the right direction. Who can quibble with the increased availability of data? After all, this can serve to improve the quality of administrative decision-making, incentivise apathetic bureaucrats to do their job, ensure course correction and, perhaps most crucially, induce transparency. What could be better for democracy?
Yet, close scrutiny reveals critical credibility challenges with the data. First, the design of data collection systems is flawed. There is no independent data collection machinery. As pointed out in The Indian Express by Kiran Bhatty and Dipa Sinha, data is collected by the very officials entrusted with implementing schemes. This creates perverse incentives for over-reporting.
And when the stakes increase, so does the incentive to misreport. There is no better example of this phenomenon than the Swachh Bharat Mission data regime. High political visibility and tight monitoring has generated reams of data, including from the National Annual Rural Sanitation Survey on latrine use and levels of open defecation (OD). But studies by the Centre for Policy Research’s Accountability Initiative reveal the tendency to over-report; local officials often report OD status based on latrine ownership rather than use, including where construction was incomplete. Importantly, an article by R.I.C.E. Institute’s researchers Payal Hathi and Nikhil Srivastava shows how training can be used as an opportunity to bias even independent surveyors to over report. In politically visible programmes where the stakes are high, these vulnerabilities need to be acknowledged, if we are to genuinely strengthen the quality of administrative data.
The tendency to misreport is exacerbated by the fact that local capacities to utilise data are weak. For all the talk of “competitive” federalism, India’s centralised administrative and fiscal structure leaves local administrators with no real power or finances to course correct in ways that genuinely improve implementation. In this context, administrative data only serves one purpose — looking good in the bureaucratic hierarchy. Finally, there are no clear data protocols or standards, which make it difficult to hold the government accountable for claims. For instance the direct benefit transfer website routinely gives a figure for monies “saved” by the government from using the scheme — a figure used routinely by politicians — but with no explanation for how it arrived at this number, rendering it meaningless.
The Modi government’s data-driven governance regime brings in full view the challenges of building credible administrative data systems in India. Importantly, it serves as a reminder of the dangers and vulnerabilities of data when they are deployed without adequate investments in quality, objectivity and credibility. Getting this right is now a challenge for the new government.
Yamini Aiyar is president and chief executive, Centre for Policy Research
The views expressed are personal