IMF stresses privacy for Aadhaar, cautious about claims of savings

Chapter 2 of IMF’s Fiscal monitor for April 2018 is about Digital Government. The document describes various efforts at digital governance across the world and makes observations about the opportunities and risks presented by them and discusses “how digitalization can change the way governments design and implement current policies, including by improving tax policy and administration, increasing spending efficiency and enhancing fiscal management”.

About Aadhaar, the IMF recognizes that it is the biggest implementation of biometrics in the world. “Biometric technology to identify and authenticate individuals can help reduce leakages and improve coverage of social programs. With more than 1.2 billion registered citizens in India’s biometric identification system, Aadhaar, the country stands out as a leader in this area.” the report says.

It makes note of the process of digitalization undertaken by India “Digitalization helped reduce leakages in two ways. First, starting in 2013, beneficiaries’ Aadhaar numbers were linked to the LPG program to prevent claims of benefits for ghost beneficiaries or multiple claims of the same benefit. Second, the government eliminated the dual pricing system and made electronic transfers of the subsidy directly to the Aadhaar-linked bank account of beneficiaries, bypassing dealers. By improving identification and verification, these reforms have reduced leakages substantially, but estimates vary. Depending on assumptions and how the reduction in leakage is expressed—that is, the reduction in total transfers or wrongful payments— estimated savings from digitalization range between 0.2 and 21 percent of cash transfers and 11 to 24 percent of wrongful payments.”

The report also avoids a common pitfall in attributing all the LPG savings to Aadhaar in the section “Lessons from Country experience”. “It is difficult to disentangle the effect of digitalization from broader macroeconomic and policy developments. For example, the use of Aadhaar in the LPG subsidy scheme coincided with the termination of the LPG dual pricing system and the reduction in the world price of natural gas, both of which helped reduce the cost of LPG subsidies. Data limitations and lack of proper assessment frameworks constrain ex-post evaluations.”

“In India, privacy and security concerns led to alternating periods of mandatory and nonmandatory use of Aadhaar in social programs. A court decision is still pending on its compliance with the right to privacy. In a recent data breach in India, it has been reported that 135 million Aadhaar numbers were compromised, underscoring the importance of sound privacy measures.” the report says, recognizing privacy issues with the project and the ongoing challenge in the Supreme Court and the at times self-contradictory roll out of the project.

The Annexure on “Digitalization of Public Finances: Country case studies” summarizes digitalization efforts in Estonia, India, and Kenya that illustrate the experiences with digitalization across different income country groups. The section on reform efforts in India acknowledges several digitalization efforts including “JAM” (Jan Dhan, Aadhaar, Mobile), DBT (Direct Benefits Transfer), the linking of Aadhaar with the existing e-filing of income tax and GST (Goods and Services Tax). When it comes to assessing the impact of these measures, the report acknowledges that it is complicated. “Digitalization can reduce leakages because of the elimination of ghost and duplicate beneficiaries and the reduction of corruption. Second, it is difficult to disentangle the effect of standalone digital measures. It is debatable whether Aadhaar was the sole source of savings or whether other parallel (digital) reforms contributed as well.”

There are several instances of savings and complicating factors discussed, notably LPG, PDS, MNREGA, and the costs of generating and maintaining Aadhaar. The report also notes the various other factors in savings in these schemes that probably had a greater impact than the digitalization and differing assessments of savings by various authorities.

The report also makes note of surveys highlighting the problems of exclusion caused by Aadhaar and factors that cause them and stresses the importance of proper availability of these factors (digital infrastructure, power and internet connectivity among others) as being necessary to the implementation of Aadhaar. It recognizes the Supreme Court ruling that Aadhaar can only be mandatory when citizens owe funds to the government (such as tax payments) but not in the distribution of social benefits.

The section on India ends with recognizing privacy and security issues. While it notes the claims of Aadhaar proponents that biometrics are encrypted, it notes the reports of data breaches and concludes “privacy and security controls are therefore key when implementing large identification programs”

Other sections of the report look at international policies and international trade online and how taxes are evaded or minimized by exploiting the differences in taxation between countries. The report also looks at the future of policy-making from the perspective of digital governance.

Full report

MediaNama’s take

The report makes some assumptions that are currently being questioned on Constitutional grounds of privacy and security. For example, it fails to recognize the legal or ethical challenges and risks of surveillance inherent in biometric identity systems when applied on a nationwide scale. The necessity of biometric identification itself is not established, nor is its effectiveness for such use. The report takes a fairly narrow perspective of digitalization of government without assessing accompanying risks, such as the tools for surveillance provided by such digitalization to a totalitarian state being incompatible with democracy.

The quoted numbers are illuminating: “Depending on assumptions and how the reduction in leakage is expressed—that is, the reduction in total transfers or wrongful payments — estimated savings from digitalization range between 0.2 and 21 percent of cash transfers and 11 to 24 percent of wrongful payments.(24)” –  The footnote (24) leads to an annexure on 3 country studies, within which these numbers are detailed, together with their sources, with higher numbers coming from the PMO and lesser from CAG. About other ‘savings’ in PDS, a tart note from Professor Reetika Khera is credited with questioning the bombastic numbers from GoI. With such swings, coming to any conclusion at all is fanciful.

There is inadequate recognition of the problems caused by exclusion in terms of the human rights of citizens being denied, though exclusion is recognized in terms of questioning the claims of savings.

News reports indicate considerable potential for the Aadhaar system actually generating ghosts, which, for the purposes of subsidies will appear to be real beneficiaries. The problems with leakages are rarely at the individual level and more usually at the level of dealers. Digitalization of individual consumption and data still has the potential to bypass the actual source of the leaks to divert resources and funds to such ghosts. However, the report fails to recognize this.

The report fails to take note of far simpler and practical alternatives to plug leakages while taking the utility of digitalization for granted. For example, with cylinders being sold to specific accounts, the appointment of inspectors to conduct cylinder to household verification has been done and proved effective in the past and is far more cost-effective and efficient than the implementation of Aadhaar as a means to prevent such leakages. Every consumer has a number that is not just linked to the address, but to the valve used to connect the bottle to the stove, without which the gas can’t be used. A large part of the savings from reducing leakages in both LPG and PDS have been the result of steps taken to directly address leakage. The report recognizes this in acknowledging the difficulty of assessing the impact of digitalization but fails to acknowledge methods that are, thus, proved more effective than digitalization.

While the report recognizes barriers to digitalization in low economy nations due to lack of necessary infrastructure (power, internet…), the report’s apparent assumption that digitalization is still beneficial is inexplicable and ignores the extensive reports of exclusion caused by such implementation.

Notably, the report makes no mention of Aadhaar as a tool of potential utility in curbing international financial malpractices – this is being claimed by the UIDAI in the Supreme Court. The concept of cross-border savings is applied only to the EU, where other mechanisms are in place to measure and perhaps even keep a check on such concealments. From India, there is almost nothing, except a curb on money going to a handful of destinations, such as Mauritius and Switzerland. Other places are considered more useful for Indian tax evaders, such as the Caribbean, where some states, such as St Kitts, are better equipped to ward off attempts by Indian authorities to trace leakages.

Actually, the report fails to even mention whether Personal Income Taxes need to be a factor for managing the economy in nations such as India. It clearly mentions that PIT is only meaningful when the economy grows, but does not mention that until it does, it is a tool to create evasion and encourage anti-social financial behaviour.

India has a major problem in political party financing, perhaps much higher than in nations such as the USA, where it is stringently tracked. However, rather than dealing with it, the country has recently created a system to anonymise donations even more enthusiastically through the creation of Electoral Bonds – a digitalization initiative notably missing in the report. It can serve as one of the factors in what the report calls ‘carousel fraud’, in connection with cross-border round tripping, but which in India does not even need to actually move across borders.

The data for expenditure on the Aadhaar project appears to be wrong. The source cited for it on the UIDAI website shows the following expenditure:

Which amounts to a total expenditure of 9942.22 crore or 99.42 billion rupees and not 87.9 billion as stated in the report. This may perhaps be due to the latest figures not being available at the time the report was created.

(with inputs from Vickram Crishna, J. T. D’Souza and Vinay Baindur)