“Data collection always has errors. Spelling mistakes in names, wrong dates of birth, incorrect or makeshift addresses… the list can go on. The work is also so time-consuming that we have had to outsource it. So, there are new private sub-contracting agencies that also add to large-scale errors in data collection and curation. Language is a barrier too. You are expecting all this work to be done in English. That is hard in Rajasthan. So, the pension got stuck because records couldn’t be mapped. Many people didn’t get their pension at all. Among those who did receive it in their bank accounts, many weren’t collecting it. We got our orders. We had to locate pensioners who weren’t collecting pension and physically verify that they are alive.
I agree that a lot of pensioners were wrongly declared dead. At times, door-to-door verification was not done at all. But, physically verifying millions of people is not an easy task. It’s not like we could hire more people to do this. It would have just cost more money and we would still have errors. We were just using local ground staff who also have other responsibilities to do this work. You have to travel through a lot of villages. Even if you do go there, there is no guarantee that the pensioner is there on that day. Many of the staff members don’t even have a vehicle. They took lifts from someone to get to the villages. They asked around about these pensioners. Even from neighbors when they couldn’t meet the pensioner directly. They had to determine whether the pensioner is alive or not. When they couldn’t get any information about them, most chose either the pensioners were dead or they were duplicate.
It is terrible what happened. It is also just a typical example of how things go wrong during data collection. Exercising judgments about data records isn’t easy. Once you are deep into a dataset, it’s hard to imagine a reality outside it. In our case, we were verifying whether a data record truly and uniquely represented someone alive or not. We were looking to confirm the connection between pensioners and their data records. When this connection is difficult to establish, it feels safer to make the record defunct. I mean, it is difficult to declare a person dead if they are sitting in front of you. Changing a column in a data record is easier. While we take responsibility for what happened, citizens should be vigilant too. Are their records correct at different government departments? We are moving towards using databases to streamline government services. We cannot do all of this work ourselves. Citizens are equal stakeholders in how we maintain their records.”
– Respondent 134, Block Development Officer involved in supervising the work of re-verifying pensioner records (personal communication, 19 November 2017).
The story above is a fictional oral account of a peculiar form of marginalization that happens through mundane bureaucratic re-categorization of citizen data records in India. In 2010, the Indian government began collecting fingerprints, iris scans, and facial photographs of every resident along with their basic demographic information—name, age, gender, and address—to give them a unique 12-digit Aadhaar (translation: foundation) number. The number is posited as a resource for state agencies to seamlessly and efficiently include registered citizens in welfare programs. This account, set in the Indian state of Rajasthan, is fictional interview transcripts in the aftermath of a bureaucratic change from money orders to Aadhaar-enabled bank transfers in the delivery of welfare pensions for the elderly in October 2015. Pension of ₹ 750 [~$11] per month is given to people of 75 and above age in India. After the change, several pensions went uncollected. In March 2016, the state government began efforts to ascertain why pensioners weren’t collecting their pensions. The government revoked pensions of about a million pensioners. Primary reasons for revoking pension were either that the pensioner was dead or their record was a duplicate database entry. When Right to Information activists held public hearings to confirm these reasons, they found many pensioners who were alive but declared dead in the database.
This fictional transcript illustrates my understanding of the distance enacted in using biometrics to uniquely identify a citizen as a data record. A data record has a life of its own. Bureaucratic re-categorization of their data as defunct records severs the relationship between citizens and their record despite the one-to-one correspondence between them. In my ethnographic fieldwork, I explore how citizens leverage their new biometrics-based relationship with the government to claim citizenship entitlements and the emergent challenges of governance through data. This fictional whisper from the field describes these challenges: it explores the affect of the everyday practice of data collection highlighting the moment of making a judgement to choose a legitimate reason from a dropdown menu for data entry on a pensioner. It is a part of a collection of three stories along with ‘The Living Dead‘ and ‘Whose fault is it anyway?‘. Together, they provide my sense of how citizens, bureaucrats, and designers of Aadhaar’s identification infrastructure rationalize such failures.
Source: Ranjit Singh, ‘Action at a Distance’, in Annie Sheng (ed.), Whispers from the Field: Ethnographic Poetry and Creative Prose (Ithaca: Cornell Council for the Arts, 2018), pp. 32-34.