And lastly, not only Article 20(3), but an overwhelming part of Indian criminal law remains dominated by the crime control model. As Chandra and Satish draw out in some detail, the remaining clauses of Article 20, the right to fair trial, the right to counsel, and the rights upon arrest and bail (which were also discussed in the previous chapter) have all been shaped by the imperatives of crime control, with due process being virtually obliterated. This is demonstrated, for example, by the Court’s readiness to uphold reverse-onus clauses in special laws, dispense with the right to counsel for pre-trial proceedings, and even question the beyond-reasonable-doubt standard for guilt in criminal law.97
One Selvi can hardly shift the accumulated weight of a few decades. But it can, at least, remind us of one thing. Society despises those who are accused of crimes. It is the easiest thing in the world for a constitutional court to strip the despised, the hated, and the shunned of their rights, claiming that it is responding to the collective conscience or the collective cry of society. It is easy to say that the rights of the individual must give way to the greater good of society, especially when society has already decided that that individual is dispensable. That is the easy way to popular acclaim and public applause.
But our Constitution exists because the easy path is not always the right path, and the Courts exist because, unlike an individual accused, they have the institutional resilience to withstand popular opprobrium when they decide to forsake the easy path for the right path. Ambedkar’s relief that the Draft Constitution had adopted the individual as its unit would end up as little more than a cruel joke if the individual has to leave behind her rights at the entrance of the police station. Selvi’s radical insight is that it is the moral orphans among us—despised by society and at the mercy of the State—who, far from being dispensable, stand most in need of constitutional sanctuary. Until that insight percolates through the rest of our criminal law, our Constitution’s transformative potential will remain unrealized.
Epilogue
‘Something of Freedom Is Yet to Come’: The Aadhaar Case, Technological Self-Determination and the Future of Transformative Constitutionalism
‘Imagine a guy walking on the road at 2 a.m., who is looking suspicious. A police patrol can take the suspect’s photograph with our app, and within a second receive details about his crime history.’
V. Gnanadesikan, CEO of FaceTagr, in an interview to FactorDaily (2018)
‘The Constitution is not a charter of servitude.’
Shyam Divan, oral arguments in the Aadhaar case (2018)
ON THE AFTERNOON OF 10 May 2018, the Attorney-General for India rose and informed the courtroom that the second-longest hearing in the history of the Supreme Court had just concluded. Spanning thirty-eight days of argument, spread out over four months,1 the ‘Aadhaar case’ had been historical in many ways. More than six years in the making (the first petition had been filed in 2012), it had generated an extraordinary amount of interest and passion both within and outside the courtroom.
Answers to parliamentary questions, World Bank reports, scientific diagrams, affidavits from rural India, and an unprecedented PowerPoint presentation had all formed part of the arguments2 in the Court. And outside it, duelling newspaper columns had carried on an intense and often acrimonious debate, which occasionally spilt over into the courtroom. Accusations of fudged documents, and uncharacteristically fraught exchanges between counsel and the Bench, had raised the temperature to fever pitch. Overall, when the Constitution Bench reserved judgement on 10 May 2018, nobody was left in any doubt about the complexity, the global significance, and the constitutional importance of the issues that the five Supreme Court justices would have to grapple with.
What was the Aadhaar case about? Pared down to the essentials, Aadhaar is a numerical identifier issued to an individual by the government. Upon enrolling for an Aadhaar number, the individual must part with her biometrics (fingerprint and iris) and certain demographic details. The biometric information is stored in a centralized database. At the heart of the Aadhaar system is the principle of authentication: information (for example, fingerprints) submitted by individuals seeking to prove their identity in order to establish eligibility for a benefit or service is ‘matched’ with the information already stored in the database, which was given at the time of enrolment. If the centralized database sends back an affirmative response, the users are deemed to have successfully proven that they are, in the words of the Unique Identification Authority of India (UIDAI),‘who they claim they are’.3
When Aadhaar was first introduced at the turn of 2010, enrolment was advertised as voluntary, aimed at facilitating citizens’ access to services by providing a swift and accurate method of authenticating one’s identity. But by the time the Supreme Court came to hear the case in 2018, Aadhaar had become virtually indispensable for carrying on life in India. It had been made mandatory for opening a bank account, for obtaining a mobile SIM card, and for paying income tax.
Biometric authentication had also been made mandatory for availing a host of basic entitlements designed for the most vulnerable members of society, such as access to the public distribution system. Private corporations, too, had begun to use Aadhaar as a mandatory means of identification. And so, the Aadhaar challenge evolved from just another run-of-the-mill constitutional challenge into something far more urgent and significant. Eventually, when it was argued, the case had become about the legitimacy of a State-created nation-wide information database, the circumstances under which it had been built, the conditions under which it was maintained, and the situations in which it could be used.
‘The Constitution is not a charter of servitude,’ Shyam Divan, lead counsel for the petitioners, memorably proclaimed during the course of arguments. Indeed, as the Aadhaar case unfolded, it became increasingly clear that it was being fought upon the terrain of transformative constitutionalism, in all its diverse, layered, and bitterly contested dimensions. At its heart, the petitioners argued, the case was about the relationship between the individual and the State, under a democratic and republican Constitution. By building a nationwide centralized biometric database, and by giving itself the power to track people’s daily transactions every time they were required to authenticate themselves, the State was distorting the balance of power that was at the heart of the constitutional order, and threatening freedom.
Elements of transformative constitutionalism were also present in the specific challenges mounted by the petitioners. By compulsorily fingerprinting almost the entire population just to prevent the possibility of identity fraud, was the Aadhaar programme not simply replicating the ‘jurisdiction of suspicion’ that had formed the basis of the repealed Criminal Tribes Act and the disgraced ADM Jabalpur judgement? Was not the forced taking of fingerprints and iris scans a severe and unjustified violation of personal privacy? Would not the very existence of a centralized database and the ability to profile individuals cast a chilling effect upon the exercise of fundamental freedoms?
Did not evidence show that the failure of biometric authentication (an inexact process at the best of times) had disproportionately impacted the most vulnerable? Had it not hit the labourer whose fingerprints had been erased by decades of manual work, the leprosy patient who didn’t have fingerprints, and the bedridden widow who could not march to the ration shop to authenticate herself? Was not this perpetuation of existing disadvantage a negation of the constitutional promise of substantive equality? By vesting unaccountable power in the hands of a new class of middlemen—those who operate the authentication machines, and therefore control the fate of welfare recipients—did not this system create a new private regime of power, flying in the face of constitutional fraternity? Was it constitutional to allow private parties into the ‘Aadhaar ecosystem’ in the absence of a data protection law? All these issues, and more, were raised by the petitioners.
On the other side, the State invoked transformative constitutionalism as well. Rampant identity fraud,
it argued, was impeding its welfare obligations under the Directive Principles of State Policy. As the most efficient and effective method of identification, biometric authentication would ensure that welfare benefits were only going to those entitled to them, and save the State substantial amounts of money by preventing ‘leakage’. Aadhaar, therefore, was a vehicle for the State to achieve the transformative ambition of Part IV of the Constitution. It was a goal important enough to justify compulsion, and the statutory framework had enough safeguards to contain the possible abuses of power and system failures that the petitioners feared. Preventing identity fraud, and thereby tackling tax evasion, banking scams, and terrorism, was also invoked as the justification for making Aadhaar mandatory for income tax, bank accounts, and mobile phones.
By a 4-1 majority, the Supreme Court upheld the constitutionality of the Aadhaar programme, while limiting, in some important respects, its scope and use.4 This is not the place for a detailed critique of the Aadhaar judgement(s) or of Justice Chandrachud’s dissent, which accepted most of the arguments advanced above.5 What is notable, however, is the Court’s failure to engage with the real terms of the debate, and the stakes at issue. While in concrete terms, the Aadhaar case was about the constitutional validity of a centralized biometric database, what it represented was something far more fundamental. Aadhaar was the first time that the Supreme Court was asked to consider how the widespread and pervasive use of technology was resetting the relationship between the State, society, corporations, and the individual. And this was a relationship that, in many ways, was now defined and characterized by data and information. Aadhaar was the first case to raise these questions, but it will not be the last. And—apart from the necessary role of Parliament in framing laws and regulations—the future of transformative constitutionalism in the twenty-first century, I contend, will depend upon how our courts will, or will not, interpret and update the Constitution to address this inescapable reality.
Through the course of this book, I have attempted to articulate a transformative understanding of the Constitution, rooted in liberty, equality, and fraternity. These guiding principles establish the framework within which democracy can thrive and flourish. In different ways, they are each concerned with the question of power. Liberty guarantees individual freedom by constraining State power. Equality authorizes, and sometimes mandates, that State power be used to remedy historical and institutional disadvantage. And fraternity seeks to dismantle social hierarchies by equalizing private regimes of power. In the cases that we have discussed in this book, relations of power have been understood in tangible and material ways: access to public goods through reservation policies, social exclusion through the practice of excommunication, physical confinement through preventive detention, and so on. While these forms of power will, of course, remain salient, what the Aadhaar case made starkly clear was that information, power, and the Constitution are now inextricably linked. Power is being increasingly exercised through, and sometimes by, technology and, in particular, through regimes and systems of information. In the opening words of the scholar Virginia Eubanks’ recent book, Automating Inequality:
Since the dawn of the digital age, decision-making in finance, employment, politics, health, and human services has undergone revolutionary change. Forty years ago, nearly all of the major decisions that shape our lives—whether or not we are offered employment, a mortgage, insurance, credit, or a government service—were made by human beings … Today, we have ceded much of that decision-making power to sophisticated machines. Automated eligibility systems, ranking algorithms, and predictive risk models control which neighborhoods get policed, which families attain needed resources, who is short-listed for employment, and who is investigated for fraud … Digital security guards collect information about us, make inferences about our behavior, and control access to resources.6
The question that we must now urgently address is how the Constitution, through the trinity of liberty, equality, and fraternity, is supposed to constrain, shape, and use this power, in continued service of democracy.
I. Information, Power , and the Constitutional Trinity
‘Constitutional guarantees cannot be subjected to the vicissitudes of technology.’
Justice K.S. Puttaswamy v. Union of India (Justice Chandrachud, dissenting)
Let us recall the foundational arguments of this book. In Part I, we discussed equality, the first pillar of the constitutional trinity. Transformative equality is not merely about requiring impartial and even-handed treatment by the State. Nor does it limit itself to asking whether State action was motivated by malice or hostility towards an individual or a class. Rather, it recognizes that disadvantage is historical and enduring. It is the product of how social institutions are designed. Inequality manifests itself in the effects that these institutions, in their working, have upon individuals and groups. Transformative equality prohibits the State from perpetuating institutional discrimination, and, where necessary, commits it to overcoming structural disadvantage through special measures and affirmative action.
The constrained vision of formal equality would view technological intervention as an unvarnished solution for social problems. After all, technology cannot discriminate on grounds of caste or sex. An algorithm cannot be biased against minorities. Transferring decision-making powers from opaque and fallible human beings to transparent and automated information systems would ensure that the State cannot, by definition, violate Articles 14 and 15(1) of the Constitution.7 Indeed, this is part of a broader narrative about the infallibility or, at least, neutrality of technology-driven decision-making, popularly known as ‘automation bias’.8
But this narrative disintegrates upon the touchstone of transformative equality, in three distinct ways. First, it ignores the human bias that exists at the time of data collection. In August 2018, the tech magazine FactorDaily provided a chilling account of how the Chennai Police used a system called FaceTagr to surveil a busy marketplace and, by querying their existing criminal database, checked whether ‘any criminals [were] identified in the area’.9 The police had no reason to suspect any particular individual of committing any particular crime: they simply engaged in dragnet surveillance to check whether anybody who had committed a crime before (and was, therefore, in the database) happened to be in the area (effectively stigmatizing anyone who had been convicted and already served their time).
The CEO of FaceTagr proudly told FactorDaily that with their app, ‘a guy walking on the road at 2 a.m., who is looking suspicious’ could be photographed by the police, and his identity cross-checked with the criminal database. This, it should be obvious, is simply the Beggary Act updated for the information age. Apart from the obvious privacy concerns, it is abundantly clear that, in the eyes of the police, ‘a guy’ who ‘looks suspicious’ for walking on the road at 2 a.m. will invariably come from a certain socio-economic class, measured simply by his appearance. And in this way, the FaceTagr app will invariably facilitate targeting the homeless and the destitute.10
Second, and of more direct concern to transformative equality, techno-utopianism fails to assess technological systems as social institutions, where discrimination and bias occur not through individual actions, but assume an institutional form.11 Even when there is no intentional bias at the time of collection, the process of collection will invariably imbibe existing patterns and structures of discrimination, which will then be fed into the model that constitutes the algorithmic system. This, by now, is a demonstrated problem and, as we can obviously see, it implicates transformative constitutionalism in its deepest sense.
Take, for example, DNA data banks, which exist in many countries, and will probably exist in India by the time this book is in your hands.12 India’s proposed DNA profiling bill provides for the storage and retention of the DNA profiles of ‘suspects’, ‘undertrials’, and ‘offenders’. A suspect’s profile must be expunged after the Police Report is filed. However, an undertrial’s profile is retained until
a court orders otherwise, and there are no provisions for expunging an offender’s profile (no matter what the character or gravity of the offence).
Effectively, therefore, the DNA Bill creates an enduring data bank of undertrials, and a permanent data bank of all offenders, making no distinction between house trespassers and murderers. These criteria appear to be neutral. Consider, however, that, whether as undertrials or as convicts, Dalits and Muslims are disproportionately represented in India’s prison population, a statistic that is linked to the fact that criminal justice regimes are inevitably skewed along socio-economic parameters.13 Consequently, if the DNA data bank ‘is primarily composed of those who have been touched by the criminal justice system and that system engages in practices that routinely and disproportionately target minority groups, there will be an obvious skew or bias in the database and the repositories’.14 Further, the retention of DNA profiles perpetuates the disadvantage that would otherwise end with the completion of the prison sentence.15
The justification given by the framers of the Bill, of course, is that storing an offender’s DNA profile will make it easier to identify him if he were to commit another crime. This logic implies, however, that once a convicted criminal, always a potential criminal. It goes far beyond even the stringent provisions of the Habitual Offenders Act, which require fingerprinting after the third conviction within a period of five years, and then require the deletion of records after a period of five more years. And its operation, as we have seen, not only tracks the institutional biases that are already constitutive of the criminal justice system, but also systematizes and perpetuates them.
The Transformative Constitution Page 38