The politics of data-based policymaking

HHow should India approach data and statistics collection in the next five years? Measures and datasets tell the stories of a country’s growth and are important for citizens, policymakers and the international public. In recent years, the accuracy and quality of Indian statistical systems have been questioned. Economists and policymakers have lamented the “decade of no data” since the Census was not conducted and the release of National Sample Survey (NSS) data was delayed. They have called for greater independence and neutrality in data collection, interpretation and release.

Are figures objective?

However, simply standing up for values ​​such as independence and neutrality ignores the contextual and political nature of statistics. We must recognize that the construction of public statistics is based on political commitments that must be expressed and criticized for good policymaking. Unless politics is transparent, the statistics will not be as useful as they could be.

Statistics are primarily attempts at numerical representations of complex social experiences. Choices about what should be counted are at the same time choices about what should not be counted. Consider welfare policies such as the Jan Dhan Yojana financial inclusion programme. India broke the Guinness World Record in 2014 for opening the maximum number of bank accounts in a week. At what cost was this speed achieved? A leader of an activist group based in Hyderabad explained that a bank had set up camps in Adivasi villages of Andhra Pradesh, hastily opened accounts for citizens and left. “Accounts were opened, but people did not know the details of their accounts and never received passbooks,” he said. Although direct benefit transfers were sent to these accounts, they were not accessible. Cases of the poor not being able to access their bank accounts and social benefits due to illiteracy, documentation issues and harassment abound, despite the statistical narrative that the number of bank accounts is increasing. In another example, the kilos of food grains distributed under the Pradhan Mantri Garib Kalyan Yojana are widely displayed on hoardings. The 2023 Global Hunger Report notes that India’s ranking has fallen from 55th in 2014 to 111th in 2023. By fixating on the number of bank accounts and kilos of food grains distributed as a measure of the state’s success, can hide the dark story of limited access to banks and poor nutrition. .

Secondly, achieving statistical objectives should not be confused with achieving development objectives. This is an issue that concerns all multilateral institutions, but India must take a more nuanced approach. Take eShram’s unorganized worker database, set up by the Ministry of Labor and Employment after the COVID-19 pandemic to collect data on migrant workers. An employee at the Common Service Center in Gurugram explained how the database relies on self-declaration, and that many who did not qualify for the employee database, such as housewives, teachers and farmers, had signed up in the hope of future benefits. While eShram has quickly met enrollment goals and has been hailed as successful, the emphasis on data obscures whether eShram is reaching its target audience. Documenting anecdotes, conducting audits and collecting experience feedback from citizens are crucial for making good development policy.

Third, more precise examination of data has become more difficult due to the digitalization of governance. For most of the 20th century, collecting socio-economic data in India was the work of public institutions, including the National Sample Survey Organization and the Central Statistical Office, which had robust details of survey methods. With Aadhaar, the state has more data about citizens than ever before. Despite the ability to anonymize and aggregate for privacy protection, it has become more difficult for citizens and researchers to access this data. Most eGovernment data is stored in state data centers that government departments and their private partners have access to, and data is published online irregularly. On the other hand, data collected through payment apps such as Google Pay and PhonePe is being used by FinTech startups to create financial products that they can sell to citizens. Data collected on citizens is not available to citizens or journalists to hold institutions accountable; only government and private actors have access.

Strengthening data systems

What are the implications of this data policy and how can we strengthen our data systems? Without denigrating quantification, we can create data systems that better serve citizens. A first step involves changing our orientation from whether we have the ‘right’ data or technical methods to asking what kind of data is most useful for the well-being of citizens. For example, while statistics on the number of newly opened bank accounts are likely to be correct, it may be useful to measure the share of poor people who have access to their bank accounts.

Second, digitally collected data should not be solely intended for use by the government and startups. It is critical to establish porous institutional structures that give civil society a voice in data infrastructure design.

Third, policymakers should expand their view of data collection from a purely technical exercise to a social and political enterprise that could benefit from input from social scientists, citizens, and activists. Ultimately, statistics should serve citizens; citizens should not be at the service of achieving statistical objectives.

Pariroo Rattan, PhD candidate at Harvard University and a non-resident research associate at the Center for Social and Economic Progress. Opinions expressed are personal