Government officials in India have often boasted that Aadhaar, the controversial national biometric identity programme, has saved the country billions of dollars by eliminating fraudulent beneficiaries from welfare schemes. To support their point, many – most recently Finance Minister Arun Jaitley – have cited a 2016 World Bank report (pdf) that estimated potential yearly savings from Aadhaar to be $11 billion (Rs 77,000 crore) per year.

In response to the government’s liberal use of the World Bank number – most significantly, in an affidavit submitted to the Supreme Court during a constitutional challenge to the Aadhaar programme – many experts have scrutinised the report and argued that the savings figure was based on shaky methodology at best.

Now, emails between one of these experts and a senior World Bank official reveal that the international body was privately critical of the Indian government’s execution of the digital identity programme, even as public statements belied this. The official says that a lack of state accountability was part of the reason why the potential savings from Aadhaar had not actually been achieved.

The emails, which Quartz has seen, are between cybersecurity researcher Anand Venkatanarayanan, who has done extensive work on Aadhaar, and a senior World Bank official who helped to produce the 2016 report.

Venkatanarayanan, who wrote an October 2017 piece in The Wire about the savings figure, had emailed the official in September that year to seek clarification on many points, including the World Bank’s stance. The World Bank official did not respond then, but when Venkatanarayanan followed up with additional queries on February 9, 2018, the person responded that same day, noting:

What you need to understand is that, the number came with three caveats that you continue to ignore: (i) it’s a potential, and not actual, estimate; (ii) it is based on extrapolation of existing studies done by academics and researchers and not by World Bank staff; and (iii) it is conditional on having an accountable government that is keen to empower and not control its citizens. In today’s India, many of the underlying conditions are not met and hence the potential has yet to be realised. (emphasis added)

The official also suggested that Venkatanarayanan should be “seeking explanation from the government as to why the (World Bank) number is being used without the above caveats?”

Venkatanarayanan responded to the official on February 10, discussing details about the study as well as pointing out that in several instances, World Bank leaders had made public statements that implied that the government had already saved huge amounts of money due to Aadhaar. One such instance was a 2017 piece in which the vice-president for the Africa region said that Aadhaar “has generated enough savings to pay for its establishment within a few years”.

The official’s response to Venkatanarayanan contained a reiteration of the point about accountable government:

It is true that many senior World Bank officials continue to talk about the benefits of digital identification systems such as Aadhaar. That’s because we do believe that digital enablers such as digital ID, digital payment platforms, and big and open data, hold the potential to make development process more inclusive and efficient. But as the WDR argues, much of this potential remains unrealised because of missing complements. And one such complement is accountable government and inclusive institutions.

The official also cited several passages from the 2016 report that showed how it is “not an unqualified celebration of technology”. One such passage reads:

The internet will thus often reinforce rather than replace existing accountability relationships between governments and citizens, including giving governments more capacity for surveillance and control.

This is very similar to much of the criticism levelled against Aadhaar.

The entire World Bank report, the official continued, “is about why despite rapid spread of digital technologies, digital dividends – growth, jobs and service delivery – have lagged behind. It’s precisely why we didn’t take your [Venkatanarayanan’s] initial criticism seriously – it is obvious that many of the complements don’t exist to make such a large saving possible at this stage in India.”

Statements such as the Africa vice-president’s praise of Aadhaar, the official clarified, “are general, subjective statements without a precise number and hence are not facts that can be disputed.” However, the official added, “I agree with you that, given the current debate in India, some caution is warranted.”

A contentious footnote

Venkatanarayanan’s email exchanges with the World Bank official also discuss the factual claims made in the 2016 report, including a highly controversial footnote.

In fact, in the initial version of the report, the footnote to the $11 billion savings claim had been attributed to a study that made no such claim. The cited study, by the Consultative Group to Assist the Poor, simply stated that $11 billion was a total estimate of the government of India’s annual expenditure on “major cash transfers”. Economists Reetika Khera and Jean Drèze have written about this footnote discrepancy in detail, as well as contacted the World Bank to clarify the matter.

After their intervention, the footnote has been changed, but the logic of the new citations do not satisfy Khera and Drèze either. In fact, the studies referred to in the new footnote cite reductions observed in two specific government programmes, and extrapolate the savings observed there to “all government of India welfare programmes”, which cost between $70 billion and $100 billion.

In a commentary piece for the Economic Times that Khera and Drèze wrote after seeking clarification from the World Bank, they point out that this capacious estimate “would include expenditure on roads, bridges, horticulture, cattle, dairy and even wildlife” – all types of expenditure that do not even pertain to Aadhaar.

“It was bad enough that when the ‘error’ in the World Bank’s savings figures was pointed out, instead of correcting their mistake, they made another convoluted attempt to justify the earlier error,” Khera told Quartz. “What is worse is that almost a year after the discredited savings figures have been exposed, the finance minister has used them again in his blog on Aadhaar.”

The official’s response states that the World Bank is “discussing the need for putting out a press release”. Almost a year later, no such public statement has been made.

The Economic Times piece was published on February 7, 2018, just before the World Bank official first replied to Venkatanarayanan. In his email to the official, Venkatanarayanan asked if the World Bank was considering issuing a public statement in order to “clear the air” about the confusion and need for clarification on the point of savings.

The official’s response states that the World Bank is “discussing the need for putting out a press release”. Almost a year later, no such public statement has been made.

Quartz is publishing quotes from these emails in public interest, especially given the continued citation of World Bank figures by the government.

When Quartz contacted the official who corresponded with Venkatanarayanan, the person said: “The information you are referring to is from a World Bank report, prepared in 2015 and disseminated in 2016. Things on the ground have changed considerably since then and we haven’t undertaken any fresh exercise in the past 4 years to make any meaningful comments on the current situation.”

The person did not comment on the ongoing situation in India, instead redirecting us to the World Bank Delhi office.

This article will be updated as and when the World Bank Delhi office responds.

This article first appeared on Quartz.