Strength in Numbers

This may not be the most topical issue to be blogging about presently, especially in the light of the general tenor of the posts before this, but I wanted to consider, for a moment, the policy of referring cases that needed to be reconsidered to a ‘larger bench’.

To put things into perspective for the uninitiated, the Chief Justices of our courts, and most visibly the Chief Justice of India, often refer cases to a ‘larger bench’ in exercise of their administrative powers. This is perhaps done under a belief that a bench of, say, eleven judges can more legitimately decide whether a previous bench of, say, nine judges were incorrect in their decision – but is this practice technically indispensable?

Unlike the constitutional court in Germany (which convenes in two panels or ‘senates’ of 8 justices each) and the United States Supreme Court (which convenes en banc, i.e. as a plenary body of nine justices), the Supreme Court of India convenes in several ‘benches’ usually of two judges. However, in certain circumstances, the Chief Justice of the Supreme Court constitutes a bench of three, five, seven, nine, eleven and in rare cases of thirteen judges to decide a case.

What are the benefits of referring cases to a ‘larger bench’? First, when two judges can’t agree upon the outcome of a case, they can resolve the deadlock by referring the case to a larger bench (usually of three judges). [Indian judicial benches do not usually convene in even numbers except in two judge benches, or in exceptionally rare circumstances where one judge in an odd numbered bench has to recuse himself from a case that otherwise has to be urgently decided.

Second, the Supreme Court of India, unlike various constitutional courts worldwide, has ‘generalist’ powers of adjudication. In other words, while the constitutional court of Germany, for example, can only decide cases involving the German constitution (‘basic law’), and the United States Supreme Court can only decide cases involving either constitutional issues or the interpretation of ordinary federal (but not state) statutes, the Supreme Court of India can (i) interpret both state and federal laws; and (ii) decide both ordinary and constitutional cases. As a result, a host of proceedings in the Indian Supreme Court are what commentators would term ‘routine’ cases. It therefore makes sense to ensure that the judges deciding the ‘routine’ cases are able to turnover cases quickly, and that they do not spend their time hearing ‘important’ constitutional issues which are likely to consume time. Accordingly, a separate bench which exclusively dedicates its time to ‘important’ cases can create a healthy division of labor.

Third, the referral of ‘important’ cases to a larger bench also follows the two-heads-are-better-than-one rationale: the principle being that if it’s an ‘important’ case, then more judges should decide its outcome.

But in order to overrule a prior case, is it absolutely essential that the subsequent case be decided by a bench of greater strength than that in the prior case?

Assume for a moment that a bench of five judges unanimously (5-0) decides an issue in terms of Holding X. Now assume that a larger bench, say of seven judges, is constituted to consider the question of whether Holding X should be overruled. Ironically, if the seven judges decide by a fractured majority, i.e. 4-3, that Holding X should be overruled, then even though more judges had agreed in the previous case in favor of Holding X, the decision can still be overruled if four out of seven judges think so. Similarly, a 2-1 majority can overrule a unanimous two judge opinion, a 5-4 majority can overrule a unanimous seven judge opinion, and a 6-5 majority can overrule a unanimous 9 judge opinion, apparently on the strength of the fact that more judges had participated in the overruling case. Therefore, the constitution of a larger bench in such ‘overruling’ cases seems to make little sense unless the rule requires a litigant to procure a larger majority than the previous case to overrule the previous decision.

However, this practice is often said to be rooted in the need for certainty. It is often suggested that the reason a bench of larger strength should be constituted to decide such cases is that it should be harder to overrule previous cases. Stare decisis dictates that prior decisions should not be overruled frivolously. However, the experience of nations has shown us that the mere fact that the decision of a certain number of judges can be overruled by the same number of judges does not make it any easier to deviate from precedent. If anything, it would require justices to understand that they do not have the power to ride roughshod over prior cases, but that they must be more circumspect while overruling a prior case, and that they must at times defer to the wisdom of the prior majority.

Am I suggesting that a bench of three Supreme Court justices should be able to overrule a previous decision of five justices? Not necessarily – but I believe that the bench strength should not really matter. What I do believe is that it is superfluous to necessarily require the constitution of a larger bench in order to overrule a prior case (a bench of the same strength could as easily do the job).

Further, the general experience has been that the larger the bench, the more intricate the law – volumes have been written on Kesavananda Bharati, and TMA Pai had to be interpreted in Islamic Academy and PA Inamdar. Of course, we could also solve the problem of intricate law by requiring our courts to speak in one voice, much in the same way that it’s said to be done in France, but not permitting justices to write concurring or dissenting opinions seems to hit hard against the essence of our legal system. Smaller benches deciding ‘important’ cases may be a less drastic way to achieve the same result.

Remember that in 1964 the Allahabad High Court in Keshav Singh’s case convened and decided an issue en banc – a decision by all 28 judges! If, god forbid, the Supreme Court of India or another court decides a case en banc: will we have to increase the strength of a court to determine if a case should be overruled? Remember also that Justice Brennan had once famously said, “the only number that matters around here is 5” i.e. a majority.

Written by
Abhinav Chandrachud
Join the discussion

This site uses Akismet to reduce spam. Learn how your comment data is processed.

  • Dear Abhinav,
    Thanks for the stimulating first post. I agree that mere number has no strength. But what you seem to suggest is that the Bench currently hearing the matter -merely because it is currently hearing, whatever its strength – has better legitimacy to overrule a precedent. This logic is equally vulnerable than the one it seeks to replace. But I am curious to know the origin of this larger bench principle, and whether there was any debate at all on the merits of this earlier.

  • Abhinav,
    Welcome to the blog and thanks for the interesting post. Venkatesan, I don't know the exact history of this, but would love to hear from others if they do. The latest iteration of this rule seems to come up in Central Board of Dawoodi (2005) – a five judge bench headed by Chief Justice Lahoti which has some relevant cites for those interested. One thing to note is that India is certainly not unique in having benches. As the post notes Germany has two senates, which actually are each further sub-divided into three chambers. Courts from Mongolia to Costa Rica have bench systems, but most have a maximum number of justices (like Germany) that can sit to finally decide a matter. South Asia – at least India, Pakistan, and I believe Bangladesh – seems to be in the minority in using a panel system where a bench can only be overturned by a larger bench. On the topic of stare decisis as Vikram Raghavan has pointed out to me recently Bengal Immunity is the case to look at where you see the Court weighing the alternative systems of the House of Lords (which didn't overturn itself at that time) and the Privy Council (which did) and coming to the conclusion that they could overturn themselves. The question though of why they need more judges to do so isn't really discussed in that case unless I'm missing something.

  • Thanks Abhinav for this engaging post.
    On your third point regarding two heads are better than one, it may be interesting to see this principle being articulated more generally in constitutional theory. In Sunstein's recent book 'A Constitution of Many Minds' (, he draws extensively from the principle of 'wisdom of the crowds' (, although for an entirely different reason – to support his theory of judicial minimalism. For an excellent critique of Sunstein's use of the 'wisdom of the crowds' principle, it is useful to refer to Dworkin's review of the book in the New York Review of Books (

  • Thank you all for the insightful comments. Nick’s reference to the 2005 case is apposite: the issue really is: In what circumstances can a court legitimately overrule its prior decisions. Our system's institutional answer (as opposed to its substantive answer), a larger bench strength, is not entirely satisfactory. VV: the weakness in the argument proposed is identified correctly, although the broader premise I’m working with is that the answer our system seems to have come up with to the issue, viz. that a greater bench strength lends legitimacy to the overruling function, is not entirely satisfactory the way it’s made to work presently. The point you make is also somewhat the point I hope to make, which is that the present system is not drastically more satisfactory than the answer I propose. Madhav: while I still have to get my hands on Sunstein’s book, I was interested by his theories of ‘ideological amplification’ and ‘ideological dampening’ when many judges decide a case, and it would be interesting to undertake empirical studies of the impact of the size of a bench on the outcome of the decision – although an Indian justice’s ideological philosophies may not be as easily decipherable as those of American Supreme Court justices.