Supreme Court of India

Supreme Court correctly holds National Tax Tribunal unconstitutional but should ordinary law be tested against basic structure and constitutional convention?

GautamBhatia_SupremeCourtofIndiajpgIn a landmark judgment last week – Madras Bar Association v. Union of India – a Constitution Bench of the Supreme Court held unconstitutional, the National Tax Tribunal Act, 2005. The majority opinion runs into 230 pages and deals with a host of complicated statutory and constitutional provisions. Briefly, the National Tax Tribunal Act provided for the constitution of a National Tax Tribunal (“NTT”), which had the power to adjudicate appeals from various appellate tribunals (constituted under the Income Tax Act, the Customs Act, and the Central Excise Act), where such appeals involved a substantial question of law. Before this, such appeals were heard by the jurisdictional high courts. The idea behind the NTT was to create a single forum for the hearing of tax appeals, in order to provide the uniformity and consistency that was lacking in the law because of the conflicting decisions of the various high courts.

The petitioners challenged the NTT Act on four grounds: first, that the reasons for setting up the NTT were fallacious, since there were no problems of consistency in the present jurisprudence of the high courts; secondly, deciding upon a substantial question of law was a “core judicial appellate function”, which could not be transferred to a quasi-judicial authority that lacked the basic features of a superior court; thirdly, that the enabling constitutional provision, Article 323B, inserted into the Constitution via the 42nd Amendment, violated the basic features of the Constitution (such as the rule of law, the separation of powers, and the independence of the judiciary); and fourthly, certain specific provisions of the NTT Act undermined the independence of the judiciary, and were therefore liable to be struck down. In short, the Court accepted the second and fourth contentions, while upholding the validity of Article 323B.

Chief Justice R.M. Lodha and Justices Khehar, Chelameshwar, Sikri, and Nariman comprised the bench.
Chief Justice R.M. Lodha and Justices Khehar, Chelameshwar, Sikri, and Nariman comprised the bench.

Examining the line of Tribunals CasesSampath Kumar, Chandra Kumar and Union of India v. Madras Bar Association, the Court concluded that it was settled law that judicial review in general – and the High Courts’ power to exercise judicial superintendence over the courts and tribunals in their respective jurisdictions in particular – was part of the basic structure of the Constitution. While the Parliament was competent to enact a law transferring the jurisdiction of the High Court with respect to specific subjects to other courts or tribunals, what it could not do was to transfer the constitutionally-vested power of the High Courts elsewhere. The NTT Act – the Court found – transferred only jurisdiction, and not the High Courts’ powers of superintendence under Articles 226 and 227 of the Constitution. Accordingly, the Court held that the NTT Act did not violate the basic structure.

We may pause here to notice a controversial assumption that the Court only touches upon: that ordinary legislation is subject to a basic structure challenge. From the inception of the basic structure doctrine, it has been a contested issue whether that doctrine applies only to constitutional amendments, or to ordinary laws as well. For instance, in Union of India v. R. Gandhi, a judgment handed down by a coordinate bench just four years ago, it had been found that ordinary laws could be challenged only upon the touchstone of the Constitution, and not the basic structure. In holding otherwise, without referring to prior doctrine, the Court further muddied an already confused area of law.

Transfer of adjudicatory functions and guarantees of judicial independence

The Court then turned to the issue of transfer of adjudicatory functions. In what is perhaps the most interesting part of the judgment, it noted cases from across commonwealth jurisdictions – Canada, England, and the Privy Council – to hold that in constitutions based on the “Westminster model of governance”, there was a clear demarcation of functions between the executive and the judiciary, with the two being kept independent of each other. This independence was guaranteed through provisions regulating the appointment and security of the tenure of judges. Consequently, adjudicatory powers could be transferred from traditional courts only if the forums that they were being transferred to had analogous guarantees of independence. This was the only way in which the scheme of separation of powers could be maintained. Constitutionally, this proposition was supported by the L. Chandra Kumar judgment, and its insistence that the transfer of jurisdiction must be to a forum that embodies the salient features of the court from which the transfer is taking place.

Facebook Images-470x246Having established this, the Court then turned to the specific provisions of the NTT Act. It found that the establishment of a central tribunal (with no regional benches), the power of the central government to determine the constitution of benches and the transfer of members, the provision allowing “technical members” and “accountant members” to be appointed to the tribunal (to adjudicate substantial questions of law), the provision of a direct appeal to the Supreme Court (thus bypassing the jurisdictional High Courts) – all pointed to the fact that while the composition of the NTT was supposed to be on the same parameters as that of the judges of high courts (since, effectively, the NTT was acting as a substitute for the High Court), in practice, this was not so. Consequently, the Court held, in paragraph 90:

Sections 5, 6, 7, 8 and 13 of the NTT Act have been held by us (to the extent indicated hereinabove) to be illegal and unconstitutional on the basis of the parameters laid down by decisions of constitutional benches of this Court and on the basis of recognized constitutional conventions referable to constitutions framed on the Westminster model. In the absence of the aforesaid provisions which have been held to be unconstitutional, the remaining provisions have been rendered otiose and worthless, and as such, the provisions of the NTT Act, as a whole, are hereby set aside.”

 What is curious about this paragraph is its reliance upon constitutional conventions to strike down legislation. This is surely a first! Constitutional conventions are – as the term suggests – conventions, that is, practices that have acquired strong normative and binding force because of a long period of consistent adherence. To hold a convention to be legally enforceable is a category mistake – what distinguishes constitutional provisions from constitutional conventions is precisely their legal standing. As the reasoning above demonstrates, the Court had no need to resort to conventions: the reasoning in the Tribunals Cases was enough for it to hold that a transfer of jurisdiction must be to a forum with the relevant trappings of a court. In the alternative, if it was of a mind to invoke the Westminster constitutional model, it could simply have held that the independence of the judiciary was a structural feature of a Westminster Constitution (as, incidentally, was held in some of the cases that the Court cited from abroad, none of which relied upon conventions). Structural analysis, that is, inferring constitutional obligations not directly from the constitutional text, but by way of necessary implications from its overall scheme and structure – is an accepted form of constitutional adjudication, widely prevalent in the United States, and – although controversial, most famously in Kesavananda Bharati – used in India as well, on occasion (see, for instance, the Delhi High Court’s finding of the word “sex” in Article 15 including “sexual orientation”). In holding constitutional conventions to be legally enforceable, the Court has – with respect – created an entirely unnecessary minefield for the future.

In a brief and pithy concurring opinion, Justice Nariman decided the case on the narrow ground that the legislature was not permitted to divest superior courts of record from the core judicial function of deciding substantial questions of law. He relied upon Chandra Kumar, which had clearly held that tribunals are competent to hear matters where the vires of statutory provisions are questioned. However, in discharging this duty, they cannot act as substitutes for the High Courts and the Supreme Court which have, under our constitutional set-up, been specifically entrusted with such an obligation. Their function in this respect is only supplementary and all such decisions of the Tribunals will be subject to scrutiny before a Division Bench of the respective High Courts.In supplanting the scrutiny of the High Court, the Act bypassed its constitutionally guaranteed power under Article 227, as well as falling foul of Chandra Kumar.

Madras Bar Association v. Union of India is correctly decided. Its holding, that the adjudication of substantial questions of law cannot be transferred to tribunals that lack the substantive trappings of the High Court, is clearly consistent with Chandra Kumar, Article 227 of the Constitution, and the structural foundations of the Westminster Model. In simultaneously holding that ordinary laws can be subjected to a basic structure challenge, and that constitutional conventions can be used to strike down an otherwise validly enacted law, the Supreme Court has, however, ventured into uncertain territory. It remains to be seen what implications this will have for the future.

(Gautam Bhatia blogs at Indian Constitutional Law and Philosophy.)