The Kenyan Court of Appeal’s BBI Judgment – II: Understanding the Popular Initiative

Article 257 of the Kenyan Constitution is one of the most interesting constitutional provisions that I’ve seen. Titled “Amendment by Popular Initiative”, it sets out ten steps for amending the Constitution, which ostensibly begin with the collection of one million signatures of registered voters, and end with a referendum. In between, there is the involvement of the representative organs (Parliament and County Assemblies) as well as a fourth-branch institution (the Independent Electoral and Boundaries Commission, or the IEBC). If each of these ten steps is completed – with its mixture of direct and representative democracy, and the participation of independent constitutional bodies – the Constitution stands amended by Popular Initiative.

As the Constitution Amendment Bill 2020 was going down the Popular Initiative Route, Article 257 was at the heart of much of the litigation, both before the High Court, and the Court of Appeal. Out of the twenty-one thematic issues framed in paragraph 50 of Musinga (P)’s lead judgment, I counted six that were directly about the interpretation of Article 257, and a few more that were ancillary. For the sake of simplicity, I propose to analyse three issues here.

  1. While the “promoters” of the BBI initiative were Mr Dennis Waveru and the Hon. Junet Mohammed, it was strongly urged that the actual force behind the initiative were H.E. President Uhuru Kenyatta and the Hon. Raila Odinga. This then led to a mixed question of fact and law: on fact, who was actually behind the initiative; and on law, if it was the President, then does Article 257 contemplate a situation where the President sets the popular initiative process in motion?
  2. The Constitution Amendment Bill contained seventy-four proposed constitutional amendments. Does Article 257 allow for this kind of a “package deal” to be put to a referendum, or does it require each proposed amendment to be put to the People in a separate referendum?
  3. What are the standards of public participation contemplated by Article 257, and did the BBI process meet those standards?

According to the Disposition, the Court of Appeal (i) unanimously held that the President was behind the BBI initiative, and that this was unconstitutional; (ii) by a 4-3 majority, held that separate amendments need not be put to the People as individual referendum questions; and (iii) did not specifically pass any orders on the third issue. Let us now consider each in turn.

Top-Down or Bottom-Up?

On the issue of who really was behind the BBI initiative, the findings are fairly straightforward. The lead judgment of Musinga (P) traces the chronology as follows: the “handshake” between H.E. President Uhuru Kenyatta and the Hon. Raila Odinga; the establishment, by the President, and through a formal gazette notification, of the BBI Taskforce; the establishment, also by the President, and also through a formal gazette notification, of the BBI Steering Committee; and the coming-into-being of the Constitution Amendment Bill 2020 as an annexure to the Steering Committee Report. Musinga (P) therefore holds:

From the foregoing, there can be no dispute that the President was the initiator of the BBI Initiative, having established and gazetted under his hand the BBI Taskforce and the BBI Steering Committee. I am in agreement with the High Court’s finding that the Amendment Bill was an initiative of the President. (paragraph 312)

During proceedings before the Court of Appeal, Appellants had nonetheless argued that the BBI Taskforce and the BBI Steering Committee were distinct from the BBI Secretariat, a voluntary alliance of political parties, and from where the actual promoters of the initiative under 257 (Mr. Waveru and Hon. Mohammed) came from. Musinga (P) rebuts this point by noting that constitutional amendment proposals were clearly within the terms of reference of the BBI Steering Committee (paragraph 321), evincing a clear intention to start the process under Article 257. Thus:

Considering the way the Amendment Bill was developed and processed, it cannot pass muster as a popular initiative. The Bill came into being after “the President and Commander –in-Chief of the Defence Forces” appointed the BBI Taskforce which prepared a report and presented it to the President, who in turn set up the BBI Steering Committee that eventually drew up the Bill. It is however not in dispute that the BBI Steering Committee toured all the counties and received views from various stakeholders, but that cannot qualify the process as a popular initiative. There is no indication whatsoever that this was a citizen initiated move. By all means, it was an executive led and driven initiative. (paragraph 325)

We find similar analysis in the other judgments (see Nambuye JA, paragraphs 66 – 71; Okwengu JA, paragraph 136, 154 – 158; Kiage JA, pgs 101 – 107, 118 – 121; Gatembu Kairu JA, paragraph 83; Sichale JA, pgs 67 – 69; Tuiyott JA, paragraphs 66 – 78, for a particularly detailed examination of affidavits).

Musinga (P) then affirms the High Court’s historical analysis, according to which a clear distinction was drawn between amendments initiated “by the People”, and those initiated by the “political elite”, with the BBI falling clearly within the latter category. A perusal of the CKRC Report reveals that the entire purpose of what eventually become Article 257 – and its passage through multiple draft Constitutions through the 2000s – was to release the amendment process from parliamentary monopoly, and provide an avenue whereby the public could be involved, not just at the end of the process (through a referendum), but at the point of initiation.

Now, what of the argument that the President could nonetheless initiate the process “in his capacity as a private citizen”, exercising his own constitutional rights? Here, Musinga (P) affirms the High Court’s conceptual analysis, noting that the Kenyan Constitution itself recognises the distinction between representative and direct democracy; consequently, a provision that expressly contemplates the latter process cannot be usurped by representative organs (paragraph 348). Other judges agree (see Nambuye JA, paragraphs 99 – 101; Okwengu JA, paragraphs 110112, and also paragraph 152, noting that the President cannot “temporary remove his executive mantle”; Kiage JA, pg 108 – 110; Gatembu Kairu JA, paragraphs 77 – 81, Sichale JA, pgs 69 – 72; Tuiyott JA, paragraphs 49, 52 – 55).

I want to make two further points here. The first is that during oral arguments, Appellants pointed out a situation where the President’s agenda would be stymied by an opposition-dominated Parliament (the famous “veto points”, to borrow a term from American Presidentialism), leaving them no choice but to take their case to the People directly. Article 257 facilitated this. This argument is acknowledged by Kiage JA, but his response is, essentially, “too bad, the Constitution doesn’t allow for that.” I think, however, that Kiage JA does indeed answer this question, but as part of his basic structure analysis (see previous post). In his analysis of comparative constitutional history, Kiage JA notes how the “Imperial Presidency” came to dominate African constitutionalism after the first wave of decolonisation in the 1960s, and he goes on to argue that the 2010 Kenyan Constitution is a response – inter alia – to the pathologies of the Imperial Presidency. This is a crucial point, because the arguments in the context of Article 257 reveal the stakes here. As multiple Justices note, the text of Article 257 does not specifically bar the President from initiating or promoting the popular amendment process. The question then becomes, how do you resolve this constitutional silence? Now if you think of the Presidential system as set out under the 2010 Constitution as empowering the President against the kind of veto-points that one finds in the United States, then the Appellants’ arguments would be persuasive; however, if you think that the 2010 Constitution was meant – inter alia – to check the Imperial Presidency, then ambiguities and silences should be resolved in favour of veto points and against expanded Presidential power. As is clear, this debate goes right to the fundamental premises of the 2010 Constitution – its “basic structure”, one might almost say! And consequently, which side one takes on this will have implications beyond this judgment, in future litigation concerning Presidential and executive powers. None of the judgments in the Court of Appeal squarely address this point, and so, arguably, it remains open.

My second point involves a debate (of sorts) between Okwengu JA and Tuiyott JA on this point. In her analysis of the legality of the BBI Steering Committee itself, Okwengu JA appears to suggest that had the Steering Committee simply floated some proposals on constitutional amendment, and had those proposals then been taken up by ordinary citizens, the process under Article 257 may have been kicked off validly. This, however, raises a concern that I had indicated in my earlier post about the High Court judgment: even a ruling clarifying that neither the President nor any other State organ can initiate or promote the process under Article 257 will leave open the possibility of doing an end-run around the Constitution through clever use of proxies. Interestingly, this danger is specifically recognised by Tuiyott JA. In paragraph 60, Tuiyott JA notes that:

That said, the process of popular initiative must be guarded from abuse. A State actor, who is otherwise barred from initiating a popular initiative, cannot originate a proposal for amendment then hire or sponsor a citizen to formulate it into a Bill and then collect signatures in support. In that instance, the promoter will simply be a surrogate of the State actor. That will not be a truly citizen-driven initiative as it will an enterprise of the State actor. There will be occasion therefore when it will be necessary to look beyond the person who formulates the draft Bill and collects the signatures to discover the hand behind the initiative, only in this way will the true intent of the popular initiative process be protected against manipulation. (paragraph 60)

Tuiyott JA therefore spends the next eighteen paragraphs minutely examining the evidence on record, including – in particular detail – the affidavit of Mr Waveru, where he himself conceded links between the BBI Secretariat, Taskforce, and Steering Committee, to demonstrate that what was happening here was indeed “an enterprise of the State actor.” This is promising: I suspect that, given the Court of Appeal’s ruling, in the future, the use of proxies – only more subtly and cleverly than the somewhat ham-fisted attempt in this case – to circumvent Article 257 is a non-trivial possibility. In such cases, Tuiyott JA’s detailed consideration of evidence indicates how the judiciary may examine this issue.

The Referendum Questions

Recall that the High Court had held that in a proposal for amending the Constitution that goes to a referendum, the proposed amendments must be submitted as separate and distinct questions, and not as a “package deal”. There are many rationales for this, two of which are succinctly summarised by Nambuye JA (paragraph 121). First, the binary, up-down nature of referenda makes them particularly unsuitable for the simultaneous determination of multiple issues, especially where citizens may have different views on those issues. By forcing a “package deal” vote, actual public preferences are thus seriously distorted. Secondly – and relatedly – the “package deal” allows the State to throw in “sweeteners” to make undesired changes more palatable. Suppose I offer to buy you ice-cream for a week if – and only if – you allow me to whack you in the face, your affirmation of my “package deal” doesn’t actually signify that you want to be whacked in the face – especially when the two “offers” are entirely unrelated.

On this point, Musinga (P) – whose opinion appears to be controlling (but see below) – adopts a textual reading that is somewhat (in my submission) at odds with the tenor of the rest of his judgment. He notes that under Article 257, what is required to be submitted to the People is a “Bill”, not a “question” or “questions”. The modalities of how this is to be done lies exclusively within the domain of the IEBC (paragraph 398). Note, however, that the textual point is not quite as clear-cut as all that. As Nambuye JA notes – in a very clear exposition of the argument (paragraphs 121 – 132) – the Article 257 differs from, say, the Article V of the United States Constitution, which uses the plural “amendments”, while Article 257 uses the singular “an amendment.” (see also Kiage JA, pgs 175 – 176) There, is therefore, at least a plausible textual argument for the proposition that the Article 257 process refers to a single amendment and therefore, by definition, excludes omnibus bills that offer up multiple amendments.

Now, on all other points, the Justices in the Court of Appeal – including Musinga (P) – consistently hold that wherever there is textual ambiguity or silence, an interpretation that strengthens public participation is to be preferred over one that does not. In this context, it is hard to argue with Nambuye JA’s observation in paragraph 128, that:

It is further my position that, when a bill is limited to a single subject, it is easier for the public to more fully understand the impact of the enactment. It also prevents fraud upon the people by minimizing the possibility of promoters of the amendment hiding harmful proposals in the midst complex multi-subject measures that the common man might not be able to grasp and understand. (paragraph 128)

I respectfully submit, therefore, that on this issue, Musinga (P) departs from his own consistent interpretive methodology. Furthermore, and somewhat bafflingly, shortly after his analysis, he nevertheless goes on to say “that notwithstanding, it is improper to lump together 74 proposed constitutional amendments in a Bill” (paragraph 399) Evidently, therefore, Musinga (P) is entirely cognisant of the problem with “package deals” – and indeed, around sixty paragraphs before, he himself identifies the problem with this package deal:

Some of the proposed amendments are rather superfluous, and strictly speaking they ought not to have been proposed as constitutional amendments by the promoters. At best, they could only be proposed as statutory amendments but were intentionally included in the Amendment Bill and appropriate statutory amendment Bills drawn by the to act as sweeteners to coax voters into supporting the proposed constitutional amendments. (paragraph 336)

He goes on to describe these amendments (tax breaks, loan exemptions etc), and immediately after, notes:

These are definitely very good and appealing proposals, but anchoring them on the Constitution of Kenya (Amendment) Bill, 2020 that also proposed very far reaching alterations of the basic structure of our Constitution was a clever bait to entice the populace, and particularly the young registered voters, who are the majority, to support the Amendment Bill, without proper civic education on all the contents of the entire Bill. (paragraph 338)

But this “clever bait” is surely as much an end run around Article 257 as is the President standing behind the figures of Mr. Waveru and the Hon. Mohammed to initiate the BBI process! Thus, this makes Musinga (P)’s finding on the issue of separate referendum questions even more baffling. I wonder, though, if an answer is to be found in paragraph 400. There, Musinga (P) notes:

I do not therefore agree with the learned judges that what is to be subjected to the referendum is a question or questions, it is the Amendment Bill, but the people are to approve or disapprove of the Bill by answering a question or questions as framed by the IEBC and approved by Parliament. (paragraph 400)

This – I would submit – is a bit of a walk-back, because here Musinga (P) does become prescriptive about the form that the Amendment Bill should take, notwithstanding the IEBC. Now, realistically, it is hard to imagine how anyone might lump together 74 amendments into a single question, unless one takes the absolutely blatant route of listing out all the amendments and ending it with a single question: “Do you agree, yes/no?” What Musinga (P) seems to be saying here is that the referendum can be a single referendum, on a single Bill, but within that, the People should have a chance to vote on separate questions separately (this possibility of a multi-option referendum as solving the “Hobson’s Choice” at issue is indicated by Kiage JA, although he declines to make a finding on it, given that there is no Referendum Act in existence yet (Kiage JA, pg 179).

There is, however, a further issue that now arises. The Disposition notes that the High Court’s finding on multiple/single question referenda has been overruled by a 4-3 majority. The three in the minority are said to be Nambuye JA, Okwengu JA, and Kiage JA. I have referred to the views of Nambuye and Kiage JJA above, and Okwengu JA says that she agrees with Kiage JA.

Now, first of all, Tuiyott JA in his opinion does not return a finding on this issue: he says that there was no “live controversy” in the instant case, as the IEBC had not yet determined the manner and form in which it would frame the reference questions (paragraph 251). To start with, this already means that – even if we read Musinga (P)’s opinion as going against the Respondents, there is no majority for the proposition that multiple issues can be lumped together in a single bill.

Complicating matters further, Gatembu Kairu JA, in his consideration of the issue, has this to say:

…[the voter’s] choice, in my view, is rendered nugatory, inoperative, and inconsequential if the voter is called upon to vote on an omnibus draft Bill, that contains a raft of numerous, diverse, and unrelated proposed amendments to the Constitution, in this case over 70 proposals of amendments, that cut across the entire spectrum of the Constitution. (paragraph 156)

He then says:

The argument made for the respondents on the principle of unity of content or single subject matter, that Constitutional amendment through a referendum should deal with only one main issue, is one I find most attractive and persuasive. (paragraph 157)

And he then says:

Ultimately, it seems to me that to put a single binary question or multiple question is a matter to be informed by the nature of amendment proposed. It may well be that certain proposed amendments may require separate and distinct referendum questions to be framed. What in my view Article 257(10) of the Constitution does not contemplate is the submission to the people in a referendum of an omnibus amendment Bill, a hotchpot of an amendment Bill, such as the Constitution Amendment Bill in this case. (paragraph 159)

Gatembu Kairu JA thus seems to adopt a middle ground, where it is possible to have thematic amendment bills, where the several questions relate or are part of the same theme, thus leading to “unity of content”. It is easy to imagine examples: say, for instance, the restructuring of a regulatory body, where it wouldn’t make sense to treat the amendments separately. This determination, it is clear, must be on a case to case basis. Gatembu Kairu JA makes this explicit when, in his summary, he says that he is not overruling, but qualifying the High Court’s orders on this point by adding the phrase “subject to the nature of the amendment” (paragraph 197); but what is also clear is that Gatembu Kairu JA – along with the three other judges in the minority – is unambiguous on the point that the lumping together of unrelated amendments into an Omnibus Bill is not merely improper, but unconstitutional.

As, however, there is nothing in the overall Disposition on this, we will have to see what the future holds.

Public Participation

The final issue that I want to consider in this post is the scope of public participation under Article 257. Public participation as a constitutional value is one of the outstanding features of the Kenyan Constitution, explicitly set out under Article 10, and subject to interpretation in a number of judgments such as Kiambu County Government v Robert N. Gakuru. Public participation under Kenyan Constitutional law has both procedural and substantive elements: transparency, adequate time, accessibility, and so on.

The requirements of public participation are somewhat challenging to articulate in the abstract, and are therefore best understood through application. In the majority judgments, the following aspects come through: (a) that copies of the Amendment Bill were posted online only in English, despite Kiswhaili being both the national and an official language (Musinga (P), paragraph 333); that there was no indication that civic education about the amendments had been undertaken (Musinga (P), paragraph 335; Nambuye JA, paragraph 84; Okwengu JA, paragraphs 122 – 128;); that the time gap between when the Bill was published in local newspapers and its approval in County Assemblies, as well as at other stages of the process, was much too short to allow for any reasonable public participation; (Musinga (P), paragraph 339; Kiage JA, pgs 135 – 137). Importantly, a majority of the bench also holds that the burden of demonstrating public participation lies upon the State, as – in accordance with the law of evidence – it has the requisite information on that point. To this I would only add: to the extent that public participation is a constitutional value under Article 10, the argument that an amendment process (for example) is not constitutionally complete until public participation has been affirmatively demonstrated, is a powerful one. Beyond the law of evidence, thus, there is a constitutional reason why the burden of proof should be upon the State.

There is, however, something of a split in the bench when it comes to the stages at which this obligation exists, and the intensity to which it exists. This split reveals something of an internal tension within Article 257. On the one hand – as almost all the Justices note – public participation is particularly vital especially in the context of Article 257, given that it is an instance of direct democracy; on the other hand, given that Article 257 is meant to be initiated by the People, ordinary people (small-p) will invariably lack the resources that will allow them to conduct public participation at a national scale, at the stage of collection of signatures. This tension is articulated by Kiage J, when he notes that:

I must express my unease, brought out quite poignantly by Mr. Karori in his address to us, that it would be to place an onerous, and well-nigh impossible burden on promoters of a constitutional amendment by popular initiative, to expect them to go the whole hog captured in the above excerpt before they can properly collect the signatures. It seems to me, with respect, that the requirements stated by the judges must be present before or as at the time the voters finally make their decision on the proposed amendments at the referendum failing which the mandatory requirement for public participation will not have been met, with fatal consequences to the proposed amendment. I am of the view, however, that the elements of public participation stated must per force be understood to form a spectrum or a continuum which is incremental in character. (pg 130; see also Tuiyott JA, paragraph 209).

Similarly, Gatembu Kairu JA notes that:

For it might appear that by one hand, Wanjiku is given a vehicle by the Constitution to propose amendments to the Constitution, but the vehicle is then taken away by the other hand, by making it impossible for Wanjiku to drive that vehicle by reason of want of resources. (paragraph 88)

Gatembu Kairu JA, however, proposes a different solution. While Kiage J would simply hold that the requirement of public participation – in all its rigour – applies at the final (referendum) stage, Gatembu Kairu JA holds that in an individual case, it would be “open” to Wanjiku to claim a lack of resources (a claim that would not be open to the IEBC, when it gets involved in the process).

I respectfully submit that Gatembu Kairu JA is correct. It is true that the burden of public participation is one that should be borne by the State – or State organs – and not by wanjiku. However, this is where we come right back to the elegant design of Article 257, which accommodates the involvement of the People, of County Assemblies, of Parliament, and of the IEBC. Consequently, would it not make sense for the requirement of public participation – in all its rigour – to be applicable corresponding to the stage at which State organs get involved? This, in my respectful submission, would resolve the tension within Article 257.


The structure and design of Article 257 gives rise to a range of fascinating questions, many of which came to be answered by the Court of Appeal. These include the difference between top-down and bottom-up amendment, with the Court affirming that, viewed in its history and context, Article 257 precludes an executive-driven process. These also include the manner in which a popular initiative may be crystallised into a referendum question (or questions), with a clear majority of the Court holding that “omnibus Bills” that have no “unity of content” are outside the scope of Article 257. And these further include the extent to which the obligation of public participation applies to a public-initiated directly democratic process. Here, the Court affirms that it is clear that in this case – given that it was executive-driven in any event – the obligation has not been discharged; it remains open, however, what standards will apply when – in the hypothetical future – a 257 process is genuinely wanjiku-driven.

I think that some of the ambiguities and tensions in the judgment(s) reflect the challenges of interpreting what is, undoubtedly, a complex provision, and also the fact that these questions are coming up for the determination for the first time, in a relatively young Constitution. Faced with these challenges, the judges of the Court of Appeal, in my respectful submission, have done important, pioneering work (as have the judges in the High Court). It now remains to be seen how future benches take some of these principles forward, and build on them.

Notes from a Foreign Field: Executive Law-Making in Romania [Guest Post]

[Editor’s Note: This is a guest post by Dragoș-Alin Călin and Irina Alexe. Dragoș-Alin Călin is a judge of the Court of Appeal in Bucharest and Co-president of the Romanian Judges’ Forum Association. Irina Alexe is an associate scientific researcher within the Institute for Legal Research ‘Andrei Rădulescu’, Romanian Academy. Readers of the blog will find this essay of interest, as it deals with the phenomenon of legislation by ordinance, in Romania.]


The starting point of this article rests on a law recently adopted by the Parliament of Romania which rejected an emergency ordinance issued by the Romanian Government, more than 13 years before. The said emergency ordinance was issued in the matter of judicial organization (GEO no.131/2006 for the amendment and supplement of Law no. 508/2004 on the establishment, organization and functioning within the Public Ministry of the Directorate for Investigation of Organized Crime and Terrorism).

It is certain that by quasi-generalising the method of legislating by simple or emergency ordinances, in the last ten years, a shift in the constitutional role of the Parliament as sole law-making authority to the Government took place. Actually, the Government has become a real legislator in almost all fields, since Parliament has changed into a sort of notary public, which many times ratifies, with or without any amendments or supplements, and sometimes rejects, after some significant periods of time, the normative acts issued by the Government, in the absence of any deadlines provided for in the Romanian Constitution for completing the Parliamentary procedure.

In its Opinion no. 950/2019, the Venice Commission acknowledged that “legislation by the GEOs became a routine. Fundamental rules of the functioning of key State institutions are changed too quickly and too often, without preparation and consultations, which raises legitimate questions about the soundness of the outcomes and of the real motives behind some of those changes. The resulting legal texts are not clear. This practice weakens external checks on the Government, it is contrary to the principle of separation of powers and disturbs legal certainty”.

Certain issues related to the legislative delegation in Romania. Statistical data relating to delegated or emergency enactment during the period 2010-2019.

The legislative delegation is regulated by Article 115 of the Constitution of Romania, republished. The competence of the Parliament to legislate may be exercised by a body of the executive power, namely by the Government, when one of the two possible types of situations indicated in the text occurs: (I) the Government’s authorization by the Parliament, under a special law, to issue, within a limited period of time, any ordinances in certain fields which cannot be subject to the organic laws, respectively, (II) the exceptional, permanent constitutional empowerment of the Government, which may be materialized by issue of an emergency ordinance, not involving an express empowerment on the part of the Parliament, to legislate in certain fields, which could be also covered by organic laws, when the conditions provided for in Article 115(4) are fulfilled and none of the prohibitions indicated under Article 115(6) occurs.

Nevertheless, the examined statistical data lead to the conclusion that the opportunity of legislation by exceptional means, by the Government, has become a rule in Romania and has not been used as an exception for a long time, as there are also years in which the Government seems to have coped with an extraordinary situation described by the constitutional text once every two or three days, which, obviously, does not reflect reality.

During the reference timeframe, 60.46% of the primary law adopted in Romania was represented by emergency ordinances and simple ordinances issued by the Government, as well as by any laws relating to their approval or rejection (a number of 2559 out of 4232).

As such, the Parliament does not operate anymore as a main legislative authority, contrary to its constitutional role, and the Government becomes the main and actual legislator. The Parliament seems to play only a secondary part; however, this is exercised late, the negative record being of 15 years and 2 months, a period needed for the rejection of the GEO no.16/1999 by Law no.66/2014. The timeframe of 13 years and one month follows in the hierarchy for the rejection of the GEO no.131/2006 for the amendment and supplement of Law no. 508/2004, by Law no. 16/2020.

For more than 60 ordinances or emergency ordinances, the Parliament needed more than three years for examination, in the case of each ordinance, including normative acts entailing an immediate approval (4 years and 6 months for the adoption of Law no.9/2020 for the rejection of the Government Ordinance no.13/2015 regarding the use of certain data recorded in the registers with the passengers’ names in the cross-border cooperation for preventing and fighting against terrorist acts; 3 years and 6 months for the adoption of Law no.238/2011 for the approval of the GEO no.53/2008 on the amendment and supplement of Law no.656/2002 on the prevention and sanctioning of money laundering, as well as for establishing some actions for preventing and fighting against financing terrorist acts; 3 years and one month for the adoption of Law no.38/2013 for the approval of the GEO no.2/2010 regarding certain measures for the organization and functioning of the working apparatus of the Government and for the amendment of certain normative acts).

Also during the reference period, only 6.29% of the issued laws contain solutions of rejection of certain simple ordinances or emergency ordinances (71 of 1128), the latter containing provisions establishing permanent legal rules, and not exclusively temporary or transitory solutions, as in the case of the GEO no.131/2006, previously mentioned.

During the period 2017-2018, three amendments were adopted to the generally referred to as laws “of judiciary”. A significant part of these amendments was harshly criticized by the Venice Commission orGRECO, being extremely detrimental to judiciary. As regards these amendments to the “laws of judiciary”, given mainly the insufficiency in the regulation, the gaps, the contradictory provisions, inadequate to the needs of the judicial system, the Romanian Government issued five emergency ordinances (GEO no.77/2018; GEO no.90/2018; GEO no.92/2018; GEO no.7/2019; GEO no.12/2019).

For example,GEO no.77/2018 was adopted with intuitu personae effects, to ensure the continuity of the positions of chief inspector of Judicial Inspection. From the publication date (5 September 2018) and until the date hereof, the GEO no.77/2018 was not approved by the Parliament and nor did the Constitutional Court of Romania rule on those three pleas of unconstitutionality raised on the dockets of courts (the oldest having been raised by a court order delivered on 19 December 2018).


Since the year 2012, the Venice Commission warned about such a constitutional situation, in Opinion no.685/17 December 2012. Almost eight years later, things seem to be identical or even more complicated, the law making in waves carried out by the Government has affected the quality of legislation and has overlooked the principle of balance of powers, and the executive power has carried out an essential and continuous role in legislating.

By the Decision no.28 dated 29 January 2020, the Constitutional Court of Romania itself notices such imbalance, considering that “the Government decision to undertake responsibility does not reflect any emergency to regulate in a given field, does not represent a measure taken in extremis, but rather taking an option opportunistic in nature, to extend over time certain temporary measures ordered under the Government Emergency Ordinance no.7/2019 and to remove from enforcement of the provisions of Law no.242/2018. At the same time, it appears that it was not necessary to adopt such a measure with maximum celerity, and possibly it could have been adopted only under the conditions in which “the major challenges” generated by Law no.242/2018 could not be overcome. 88. As regards the importance of the regulated field, it is found that, in principle, the field of justice is one in which the Government may undertake responsibility [see, for example, Decision no.375/2005]. (…) 90. Given the above, the Court acknowledges that the criticized laws breaches Article 114 of the Constitution and, implicitly, Article 61 paragraph (1) of the Constitution, by excessive limitation of the role played by the Parliament. Hence, it also appears a breach of the constitutional principle of balance of powers [Article 1 para. (4) of the Constitution] since one of the State powers, namely the executive power, assumed a preeminent role in the enactment activity, excessively using an enactment procedure which, by its nature, is exceptional.”

The significant timeframes within which the Parliament approves or rejects normative acts issued by the Government, in the absence of clear time limits provided for in the Constitution of Romania, are liable to lead to subordination, which is contrary to the principle of balance of powers.

In such a context, until an expected constitutional revision, a possible solution would be found in the principle of loyal cooperation and of mutual respect between the state authorities/institutions.

Notes From a Foreign Field: The Ugandan Constitutional Court on the Right to Protest [Guest Post]

[This is a Guest Post by Karan Gupta.]

In a society where policy brutality and clamp down on free speech is common, the Constitutional Court of Uganda recently affirmed a few commonsensical principles on free speech, the right to assemble and public order. On 26 March 2020, the Constitutional Court of Uganda declared Section 8 of the Public Order Management Act 2013 (POMA) unconstitutional (4-1 majority). Section 8, inter alia, conferred upon the Inspector General of Police (IGP), or any officer authorized by them, vast discretionary powers to: (i) Withhold permission to hold a public meeting or stop a public meeting where it is “held contrary to the Act”; (ii) Use force to disperse public meetings; and (iii) Impose criminal liability on organizers and participants of such public meeting. The Act defines a ‘public meeting’, empowers the IGP to regulate their conduct (S. 3), and requires every ‘organizer’ to give prior notice (at least three days prior and no more than fifteen days prior) of the proposed public meeting with details specified therein (S. 5). An unplanned, unscheduled and unintended public meeting is exempt (S. 7).

I explore, in seriatim, the constitutionally flawed approach, the progressive observations of the constitutional court, missed opportunities, and lessons for India.

 Preliminary Point: Constitution adjudication – in personam?

In a previous case, Muwanga Kivumbi v Attorney General, the Constitution Court had declared Section 32(2) of the Police Act 2006, which empowered the officer in charge of the police to pass an anticipatory order prohibiting the convening of an assembly/procession if there were “reasonable grounds for believing” that there would be a breach of peace, unconstitutional. The Court had held that the subjective and anticipatory power was prohibitory in nature (as compared to a regulatory power, which is permissible) and ultra vires Arts. 20(1) (Fundamental Rights are inherent and not granted by the state) and 29(1)(d) (Freedom of assembly and demonstration).

The challenge in HRNU lay in narrow confines. Art. 92 of the Constitution restricts the Parliament from passing a law which “alters” a decision of the Court “as between the parties to the decision or judgment”, thus barring the alteration of rights that have accrued to parties to a case vis-à-vis each other (in personam). The petitioners in HRNU highlighted that one petitioner in Muwanga was also a petitioner before the Court in HRNU, and that Section 8 of the POMA was para materia to Section 32(2) of the Police Act. Consequently, by enacting Section 8, the legislature had unconstitutionally attempted to alter the decision in Muwanga [p. 7]. Despite a broad challenge to the POMA on a myriad of constitutional provisions, the petitioners restricted their oral arguments to only Art. 92 and Section 8.

Justice Cheborion evaded the limited ambit of Art. 92 and the nuanced differences between the two provisions (Sections 32(2) and 8)) and held that Art. 92 also applies to decisions made in public interest, and not only in relation to parties to a previous litigation. This approach was adopted by two other judges [p. 42, 50, 69] and raises two concerns: First, this militates against the plain and ordinary meaning of Art. 92 and renders nugatory the latter part restricting its application to parties in a litigation. The decision in Muwanga (as well as the present case) concerned a constitutional matter on the ambit of constitutionally permissible police powers to control assemblies, demonstrations and peaceful protests. Such matters are, by their very nature, in rem proceedings – a declaration of invalidity does not operate only between parties, but to everyone. Art. 92 which seeks to protect rights that accrue to parties from a litigation (from contract law or property law for example), has no applicability in such cases. Furthermore, the legislature did not seek to alter the decision in Muwanga, but enact a new provision different from Section 32(2) of the Police Act.

Second, the Ugandan Parliament may alter the basis of decision in Muwanga by amending the provisions on which the decision turned [i.e. Arts. 20(1) and 29(1)(d)]. However, the Court held that the decision in Muwanga could only be altered, inter alia, by amending Art. 92 [p. 19]. This flows from its erroneous reading and application of Art. 92. The right approach is for the Court to, absent any constitutional amendment, employ in its assessment the broad constitutional principles laid down in Muwanga and other relevant constitutional provisions. However, the court restricted its assessment to whether Section 8 is an “incarnation” of Section 32(2) [p. 18]. Only two judges avoided this pitfall, though without adequate explanation [p. 54, 75], analyzed the entire Act and declared it unconstitutional [p. 61].

 Public interest, public order and free speech

Art. 29(1) guarantees to every person the freedom of speech and expression, assembly and demonstration, and association. While Art. 79(1) empowers the Parliament to enact laws for the maintenance of order, both provisions are silent on the permissible restrictions on fundamental rights. The answer is found in Art. 43 which stipulates that “no person shall prejudice the fundamental or other human rights and freedoms of others or the public interest”. According to Ugandan precedent, there is no justification to restrict or abrogate a fundamental right where the its exercise comports with the restrictions in Art. 43. Art. 43(2) clarifies that the term ‘public interest’ shall not permit any limitation “beyond what is acceptable and demonstrably justifiable in a free and democratic society”.

The scheme of fundamental rights chapter is significant for two reasons: First, barring the general restrictions in Art. 43 and a few other provisions which specify restrictions therein, there are no specified grounds to restrict fundamental rights as compared to other Constitutions; Second, under Art. 43, the principles of a free and democratic society are accorded primacy and any restriction must comport with this requirement. While the Parliament may legislate on the maintenance of public order (which Muwanga held is in public interest), any restriction on the freedom under Art. 29 must be demonstrably justifiable in a free and democratic society.

Viewed in the above context, the Court (three or more judges) made three significant observations:

 First, on the ambit and hierarchy of free speech protection, the Court held that speech, public processions and protests, irrespective of their nature, are entitled to equal protection (i.e. social, religious, political, economic, and so on) [p. 20, 57]. This is distinct from the preferred position doctrine which accords higher protection for political speech in American constitutional jurisprudence. The context which informed this analysis is granting equal protection to political assemblies and speech which are a common target of the political establishment in Uganda;

Second, on public order, the Court held that where a protest or public gathering is peaceful, “it does not matter that it may disruptive or even inconveniencing”. [p. 21] This is significant as any society committed to the freedom of speech and assembly recognizes that some disruption is no ground to restrict or deny the right. Beyond toleration, the celebration and protection of speech and assembly is linked to justice, equal concern and mutual respect of every individual. Recall here the judgment of the Madras High Court which affirmed that “public streets are the natural places for expression of opinions” (analyzed here); and

Third, the Court held that Section 8, in so far as it authorizes the police to prevent a public meeting, empowers them to impose a blanket ban and require prior permission for every gathering [p. 23, 25]. The Court held that this violates Arts. 29 and 43 and the state failed to demonstrate that the power conferred by Section 8 is both regulatory and acceptable and demonstrably justifiable in a free and democratic society. This is significant for three reasons:

  • The burden of proof to justify the restriction on a constitutionally guaranteed right falls on the state. However, by vesting in the police the vast discretion to prohibit or prevent the freedom of speech and expression, this burden stands reversed. Every individual is then required to justify to the police why the exercise of their constitutional right will not or does not impair public order or contravene any provision of POMA. Justice Elizabet rightly noted that there is a presumption that every assembly and exercise of free speech is peaceful [ 61] and a mere apprehension of violence does not constitute a sufficient basis to prevent or prohibit an assembly, gathering or protest;
  • The Court highlighted that no conditions were laid down for the exercise of the power [ 98], vesting in the police vast discretion to determine which public meetings may be prevented or forced to disperse. As noted by the American Supreme Court in Grayned v Rockford, such unbridled discretion is accompanied by the “attendant dangers of arbitrary and discriminatory application.” Where the police are an instrumentality of the state, the possibility of partisan politics to curb dissent and anti-establishment sentiments cannot be discounted. This is recognized by two judges in HRNU who documented the arbitrary exercise of power by the police to protect government interest and impose popular morality [p. 22, 100]; and
  • Flowing from above, the unregulated blanket discretion to prohibit assemblies does not comport with the requirements of the ‘constitutional yardstick’ that every restriction on a fundamental right must be necessary and The police must justify, in each specific instance, why the prohibition of an assembly is the least restrictive measure and proportionate to the possible harm sought to be prevented (this post discusses context-specificity and proportionality in the context of the internet shut-down judgment in Kashmir). The judgment, despite lacking in the explicit use of this yardstick, comports with this requirement.

Maintaining public order

How then must the State maintain public order? Justice Cheborian (with whom three other judges agree) answered this. He held that where the police anticipate a breach of peace, there is a positive obligation on the state to provide protection and police deployments and not prohibit the assembly [p. 17, 60, 90]. The Court held that the duty to maintain public order “cannot be discharged by prohibiting sections of the public from exercising their constitutionally guaranteed rights to demonstrate peacefully or hold public meetings of any nature.” This is sound as it reaffirms: (i) the holding in Muwanga that the state must provide channels and structures to ensure that legitimate protest “find voice”; and (ii) the principle that the failure of the state to provide adequate security cannot be a ground to deny people the freedom of speech and assembly. Furthermore, the Court noted that the state is also empowered to act in various situations (unlawful assemblies, riots, malicious damage against public order) by the Penal Code of Uganda [p. 33,89] and may, in accordance with the law, arrest or take appropriate action against any perpetrator [p. 23, 32].

Missed opportunities

 Despite the significant observations above, the Courts missed out on two opportunities:

First, only Justice Kenneth attempted to specify which values underlie a free and democratic societyinter alia, the acceptance and accommodation of a variety of cultural, religious and political beliefs and free political debate, human dignity and freedom of speech, association and movement [p. 90]. The Court could have laid down a comprehensive base for the protection of free speech and association as an values inhering in a free and democratic society; and

Second, despite the challenge in the petitions to numerous provisions of the Act, the Court examined the validity of only Section 8 (admittedly, only this was pressed by the petitioners). Only two judges examined the deeply inherent flaws in the entire Act to conclude that it was ultra vires a myriad of constitutional provisions. In the end, these draconian provisions were left standing and the Court concluded that guidelines must be framed for the exercise of powers under the Act [p. 33]

India and Section 144

Recall here that Section 144 of the Indian Code of Criminal Procedure 1973 confers wide discretionary powers upon executive magistrates to prohibit assemblies. This is regularly invoked on the basis of an apprehension that there would be a breach of peace or public order. The Constitution Court in Muwanga struck down a similar provision [Section 32(2) of the Police Act] on the ground that it was prohibitory in nature and reversed the burden of proof – every individual was required under it to justify why the exercise of the right to free speech and assemble would not cause a breach of public order. The Court held that this suppresses the “powerful tool” of peaceful assemblies and protests when a free and democratic society must encourage the “greatest possible freedom of expression.” This reasoning was reiterated in HRNU to strike down the wide power to impose blanket anticipatory bans on assemblies under Section 8 of the POMA.

The power under Section 144 [similar to Section 32(2) and Section 8] allows the imposition of anticipatory bans and is prohibitory in nature This falls foul the constitutional standards espoused in Uganda. Indian courts have attempted instead to narrow the discretion conferred by the provision. The judgment of the Supreme Court in Anuradha Bhasin (internet shutdown in Kashmir) recently affirmed that a valid exercise of power under Section 144 is premised on: (i) the existence of objective material facts which form the basis of the opinion formed by the Magistrate; (ii) its general invocation being confined to a specific area and issue; (iii) the existence of a demonstrably urgent situation; (iv) such measure being the least restrictive course of action; and (v) compliance with proportionality standard.

Despite these restrictions, the burden of proof continues to rest on individuals as magistrates are empowered to impose anticipatory orders prohibiting any assembly. Even where these orders are challenged before courts, the preliminary burden falls on the individual (as a petitioner) to prove that the issuance of the order violated their fundamental right. This was most evident in December 2019 where when faced with the legitimate expression of dissent against the Citizenship Amendment Act, numerous orders under Section 144 were imposed across the country without any well-founded apprehension of violence, citing inconvenience.


Despite its failings, the judgment in HRNU held that there is “absolutely no legal authority” to stop peaceful expression on the basis of an alleged breach of peace [p. 22] and builds on Muwanga to add to the growing jurisprudence that restrictions on the right to free speech and assembly are exceptions which the state is required to justify in every case prior to its imposition, reaffirming a commitment to a culture of justification, not authority. With this, it also provides guidance to India in the exercise of the power under Section 144. While Section 144 remains operative, one can only hope that the government of the day responds proactively to the protection of the right to free speech and assembly in scrutinizing more closely the impositions of these orders in the first place.

ICLP Round Table: The Equality Bill 2019 – Comparative Perspectives

(This is a guest essay, authored by Professor Sandra Fredman, who is Rhodes Professor of Law at the University of Oxford. It is the third essay in the ICLP Blog’s Round-Table on the Centre for Law and Policy Research’s Draft Equality Bill 2019. The first two essays are available here and here).

India is unusual in that the provisions guaranteeing equality in the Constitution have not yet been supplemented by comprehensive anti-discrimination legislation comparable to that in other jurisdictions, such as the US, South Africa or the UK. Anti-discrimination legislation plays an important role in giving specific content to constitutional guarantees. It can also bind individuals and corporations, thus extending the reach of the right to equality beyond the State into key sites of discrimination, such as employment, housing, education, healthcare and goods and services available to the public.  The proposed Equality Bill drafted by Jayna Kothari and her colleagues at the Centre for Law and Policy Research is thus very much to be welcomed and she and her team deserve to be congratulated on the quality and reach of the proposal. This post takes a comparative perspective on the Bill, while of course being sensitive to the very different context in which the new legislation will function.

There are many aspects of the Bill that demonstrate positive advances on relevant comparators, such as the UK Equality Act 2010 (EA) and EU legislation. This is particularly so for the definition of direct discrimination. Direct discrimination has traditionally been defined as occurring when a person is treated less favourably than another due to a protected characteristic. The need for a similarly situated comparator, however, has raised complex issues both in principle, because it requires conformity with a dominant norm as a gateway to equal treatment, and in practice, where the choice of comparator can be contested. This has most saliently occurred in relation to pregnancy, with the sad spectacle of tribunals and courts in earlier decades refusing to accord pregnant women the same treatment as men because there is no appropriate male comparator. The EA addressed this problem by replacing the requirement for a comparator with the simple need for detrimental treatment. But this only applies in relation to pregnancy (and disability in limited circumstances), The Indian Bill takes a step further and gives both possibilities in the alternative, available for all protected characteristics, not just pregnancy.

The Bill is also important in that it aims to address structural discrimination in various forms. While the EA addresses such discrimination solely through indirect discrimination, which focuses on the effects of apparently neutral practices, provisions or criteria, the Bill also includes a provision expressly defining ‘systemic or structural discrimination,’ defined as hidden or overt patterns of institutional behaviour, cultural traditions and social norms that discrimination against individuals with protected characteristics. The Bill’s definition of indirect discrimination, is however, interesting in that it does not include the usual justification defence for respondents, permitting the latter to demonstrate that the provision is proportionate to a legitimate aim (such as a requirement for a relevant qualification which might be necessary for a job even though it excludes a disproportionate number of people with a protected characteristic). 

The Bill is impressive too in the wide range of characteristics which it classifies as protected against discrimination. Although many of the listed characteristic can be found in international or regional sources, the Bill contains several important additions, including linguistic identity, migration, refugee status, occupation and food preference. The Bill also transcends the straight-jacket of the Constitutional equality provisions, which only list a set number of protected characteristics. Instead, its list is non-exhaustive, including a provision by which analogous grounds can be included if they fit the principles in the Bill. In addition, the Bill makes express provision for addressing intersectional discrimination. Notably, the Bill includes socio-economic status as a protected characteristic. While this is a welcome move, this formulation runs the risk of protecting the stronger as well as the weaker sections of society. This might open the way to claims by advantaged persons that provisions such as higher taxes for wealthier persons might be challenged as discriminatory on grounds of socio-economic status. The drafters might instead want to use an expressly asymmetric term such as socio-economic disadvantage.

One aspect of the  Bill which might also need further attention concerns the scope of the protection for workers. Given that over 80% of employed persons in India earn their living in the informal sector in India,  dwarfing the formal sector in magnitude, a key initial question is the extent to which non-standard workers are protected. This is a highly challenging issue, and there are few modern positive precedents for its resolution. In particular, most labour law systems confine eligibility for employment protection to workers working under a contract of employment, excluding self-employed, non-standard workers. Discrimination law in the UK extends to some self-employed workers, provided they work under a contract to provide a personal service; but this too has opened up a large number of loopholes for those attempting to sidestep their obligations.

The Bill certainly takes important steps towards inclusion of non-standard workers, such as including employment agencies in the prohibition on discrimination, and mentioning domestic workers in the definition of employer. However, the way in which ‘employee’ is defined might still exclude some of the most vulnerable workers outside of the formal sector. ‘Employee’ is defined as any person ‘employed on wages in any establishment, factory, company or shop.’ An Uber driver who is discriminated against because he is Muslim, a woman agricultural worker  who is sexually harassed, or an own-account worker who is paid less because of race, caste or migration status, might find it difficult to bring themselves within this definition. One way forward is to provide separate protection for self-employed workers who are discriminated against by clients or customers, similar to the protection for consumers or users of public accommodations. Another is to replace the use of ‘wages’ with earnings, and give less emphasis to the workplace as the site of work.  Also of importance is to bring the provisions for equal pay for work of equal value in the Equal Remuneration Act into closer alliance with the Bill.

The Bill is also noteworthy in its inclusion of proactive duties, closely resembling the  ‘fourth generation’ equality duties found in the EA. Such duties are important in their recognition that individual complaints of discrimination will never be sufficient to address the structural causes of discrimination. Proactive duties on the part of the State are crucial in discerning and changing such systemic patterns, as well as taking the burden off individual complainants. The Public Sector Equality Duty in the Bill tracks closely onto its UK equivalent by requiring all appropriate authorities to ‘have due regard to the need to’ eliminate prohibited conduct under the Act, remove or minimize disadvantage, and encourage participation. It goes somewhat further than the EA in that due regard should also be had to taking steps to meet the need of persons with protected characteristics. However, given the extent to which the ‘due regard’ standard under the EA has become little more than a ‘tick-box’ procedure, the drafters might consider strengthening the requirement. For example, rather than requiring authorities to ‘have due regard’ to the need to eliminate prohibited conduct, the drafters might consider requiring authorities actually to eliminate such conduct, which is prohibited in any event. Similarly, rather than imposing a duty to have due regard to the need to take steps, they  might think of imposing a duty to actually take the steps.  Nor would this be pushing the boundaries too far, given that the previous clause requires the State to take measures to promote equality.

Even more welcome would be the inclusion of a duty to have regard to the need to reduce economic inequalities. A similar duty was inserted into section of the EA, which required public authorities, when exercising strategic functions, to have regard to the desirability of reducing socio-economic inequalities. Despite being a very mild obligation, this provision has never been brought into effect, despite remaining on the statute book, so that there is no experience of its workings. For the Indian Bill, it would not be too large a step to include such a duty, given the emphasis in the preamble on eradication of socio-economic inequalities, and the inclusion of socio-economic status in the list of protected grounds. Consideration might even be given to strengthening the obligation.

All in all, this is a very welcome intervention, and it is to be hoped that it will soon find its way onto the statute books.

Notes from a Foreign Field: The Namibian Supreme Court on Free Speech, National Security, and Injunctions

In an interesting judgment delivered earlier this month (Director-General v Haufiku), the Supreme Court of Namibia restated some common-sense principles about the relationship between freedom of speech, national security, and judicial injunctions. The facts in Haufiku were straightforward: The Patriot, a Namibian newspaper, had uncovered some information about potential corruption within the Namibia Central Intelligence Service [“NCIS”]. It appeared that the NCIS had purchased farms and houses using public funds, which was then given over to a private association of ex-NCIS employees. When The Patriot’s journalist wrote to the Director of the NCIS with queries about these purchases, he was informed that his questions fell within the scope of “sensitive matters and/or classified information” (para 11). Under the terms of the Namibian Protection of Information Act [“PIA”] (whose provisions are strikingly similar to our Official Secrets Act) and the Central Intelligence Services Act [“CISA”], not only was the NCIS not obliged to provide information, but also, possession and publication of such information was an offence. Accordingly, the NCIS subsequently approached the courts for an interdict that would prevent The Patriot from publishing what information it did have.

General Grievous Meme

The State’s Arguments

It was argued before the courts that the information in The Patriot‘s hands violated the PIA, as it related to a “prohibited place” and/or a “security matter.” In particular, it was argued that “if the information were published, it would threaten or jeopardise the national security of the State” (para 18), “any disclosure of information which showed either the capability or a lack of resources on the part of the NCIS is unlawful as it undermines the effectiveness of the institution and with that posed a security vulnerability to the State of Namibia” (para 19) When asked to substantiate this argument, the NCIS argued that it was not obliged to do so, because the Courts did not have any jurisdiction to assess national security questions on their merits (paras 37 and 44 – 48). In short, the NCIS made a rather meta argument: the information that The Patriot wanted to publish impacted national security, but how it did so could not be revealed (inter alia, because that itself would be tantamount to impacting national security). As the Court characterised the submission:

  • The NCIS is the sole determiner of whether or not there is a threat to national security from the disclosure of information by a member of the public and not even the courts may inquire into that;
  • The NCIS is not obliged (in fact it is prohibited not) to place evidence before court in court proceedings to justify its conclusion that publication will be harmful to national security;
  • All the NCIS needs to do in court proceedings aimed at supressing publication of ‘secret’ information- be it about its assets or anyone associated with it – is to assign it the label of national security and to assert that publication will compromise national security and the court is bound in law to grant an interdict prohibiting publication;
  • The NCIS is under no obligation to reply to any enquiry by the media or to comment on any matter relating to or concerning the NCIS, even if it involves an allegation of a crime such as corruption. (paragraph 50)

(Readers who followed the Attorney-General for India’s arguments during the Rafale review petition will recall that this was more or less exactly the same argument advanced by the State in that case.)

The Court’s Analysis

The Court began by making the sharp observation that “the written submissions refer altogether to a staggering 50 cases a significant number of which are pre-independence cases decided under the pre-independence securocratic ethos which conjure up images of our painful colonial past.” (paragraph 43) (N.B.: “securocratic ethos” is a brilliantly evocative phrase to describe one of the fundamentals of colonial regimes all over the world!) Moving on from that, it then observed that, as a matter of legal burdens, because the NCIS was the body that was seeking the injunction (instead of a situation where, for example, the journalists were seeking disclosure), the onus was upon it to demonstrate what right was being interfered with. For this, “the mere assertion of a reasonable apprehension or fear of interference would not suffice. The facts supporting the apprehension must be set out in the application to make it possible for the court to make an assessment itself whether the fears are well grounded.” (paragraph 62)

Crucially, the Court then went on to note that this basic legal requirement – of establishing facts – did not go to sleep in a sand-box just because the State was invoking national security. As the Court made clear:

It needs to be made clear as a preliminary matter that we do not agree with the Government’s refrain, repeatedly pressed with great force in the written heads of argument, that once the Executive invoked secrecy and national security, the court is rendered powerless and must, without more, suppress publication by way of interdict.

The notion that matters of national security are beyond curial scrutiny is not consonant with the values of an open and democratic society based on the rule of law and legality. That is not to suggest that secrecy has no place in the affairs of a democratic State. If a proper case is made out for protection of secret governmental information, the courts will be duty bound to suppress publication. (paragraphs 84 – 85).

In the instance case, the Court found that neither with respect to the Association, and nor with respect to the properties that it had purchased, had the NCIS made any kind of case – or submitted any kind of evidence – to demonstrate national security concerns. For example:

…the information about the properties is not inherently secret (such as a military installation, equipment, password etc.), making it obvious to anyone who possessed it, even inadvertently, that it concerned or was a matter of national security. What is in issue are a house and a farm which are readily accessible to the public – without any indication that secret government operations were being carried on there as contemplated in s 23 of the NCISA. (para 102)

For these reasons, the Court denied the claim for injunction.

National Security


While at first blush, the Supreme Court’s judgment seems to be doing nothing more than recapitulating well-settled principles of law, there are a few reasons why, nonetheless, it is an important judgment in a comparative context. The first is that notwithstanding how well-settled these principles are, they come under repeated challenge from State authorities. As I mentioned earlier in this essay, readers will recall that the Attorney-General’s arguments in Rafale went along precisely these lines: the Court was asked to adopt a hands-off approach towards the fresh documents on national security grounds; when it was pointed out to the AG (by the Chief Justice) that under evidence law, all that mattered was a document’s relevance, and not how it had been obtained, the AG asked the Court to carve out a special exception to this rule for “national security” matters (a request that the Court fortunately declined). The AG also declined to justify his arguments, again on the basis that the moment “national security” was invoked, the Court had to back off. This is precisely the kind of “securocratic ethos” that the Namibian Supreme Court decried, and which is simply inconsistent with the principles of an open, liberal-democratic society.

The second important point is that the Namibian Supreme Court also made clear that invoking national security did not exempt the State from having to prove its case in the normal manner. While some necessary leeway could be made – such as having an in-camera proceeding while adjudicating the State’s claims (or sealed covers?) – the legal standard would remain the same. Invoking “national security”, therefore, would not cloak the State with some special kind of immunity; when it came to an injunction and the suppression of speech, the normal standards would apply. This, of course, has been the teaching of a number of Courts ever since the Pentagon Papers case (even if the Namibian Supreme Court’s judgment did not quite rise to that level, and had perhaps some unnecessary observations about hypothetical cases where it would grant injunctions). It is also important, because even if a Court does not except that it is divested of jurisdiction to adjudicate such cases, a highly deferential standard towards the State effectively achieves the same outcome. What is needed, therefore, is a shift from the “securocratic ethos” – where the invocation of national security is given a presumptive deference that is not extended to other claims – to the “constitutional ethos” where, ultimately, what is supreme is the Constitution, and constitutional values (including those of free speech).

It is in that sense that the Namibian Supreme Court’s judgment represents an important step forward in the global struggle against the permament entrenchment of the “securocratic ethos”, and towards an open society.


The Afterlife of the Aadhaar Dissent: The Jamaican Supreme Court Strikes Down a National Biometric Identification System

Justice Charles Evan Hughes’ famous aphorism, that a dissent is an “appeal to the brooding spirit of the law, to the intelligence of a future day”, has passed into legend. It was famously invoked by Justice H.R. Khanna, while concluding his dissent in the Habeas Corpus case. But sometimes, a dissent is not limited to a footnote in the judicial lore of a nation, waiting for the years to pass by until the “intelligence of a future day” dawns. Sometimes, like the swallow flying south, a dissent becomes part of the global migration of ideas. It finds fertile soil far from home, there to bloom into the full richness that it has been denied in its own native environment.

Earlier this week, the Supreme Court of Jamaica struck down the Jamaican National Identification and Registration Act [“NIRA“]. The NIRA was a law that mandated the collection of biometric information from all Jamaican residents, and its storage in a centralised database. The similarities with Aadhaar are immediate and clear, and indeed, a substantial part of the judgment(s) were focused on comparisons between NIRA and the Aadhaar Act, and the Indian Supreme Court’s judgment in Puttaswamy. And what is striking about the judgments of Chief Justice Sykes and Justice Batts is that, after a detailed and painstaking engagement with Puttaswamy, both judges held that, on multiple counts – on the application of the proportionality standard, on the articulation of freedom and privacy, on the dangers of centralised databases, and on the shape and design of data protection frameworks, it was Chandrachud J.’s dissenting opinion that was persuasive, and deserved to be followed. The NIRA, accordingly, was struck down as unconstitutional.

NIRA: Similarities and Differences with Aadhaar

As the sketchy description outlined above indicates, NIRA and Aadhaar bear substantial similarities. Indeed, these similarities went beyond the collection and storage of biometric information, and extended to the design of the enactments: like Aadhaar, NIRA established a centralised database [“NCID“], a centralised authority to deal with collection and storage, contained provisions for private party access, had provisions for when data could be accessed, and so on. Furthermore, the justifications offered by the State were remarkably similar: preventing and eliminating crime, preventing corruption (FATCA was cited as well!), streamlining delivery of benefits, facilitating ease of identification, and so on.

At the same time, however, there were some substantial formal differences between NIRA and Aadhaar (why I use the word “formal” will become evident immediately hereafter, to anyone familiar with how Aadhaar actually works in practice). First, enrolment under the NIRA was legally mandatory, on pain of criminal sanctions; secondly, information collected under NIRA was substantially more than that collected under Aadhaar (it included blood samples, for example); and thirdly, third parties had formal access to the database under NIRA.

The judgment(s) of the Jamaican Supreme Court made much of these differences between NIRA and Aadhaar to argue that criminalising non-compliance was disproportionate, that the information taken did not comply with the principle of data minimalism, and that the access of third parties raised constitutional concerns of the privacy and security of the data. What is remarkable, however, is that despite finding these substantial formal differences, and despite finding that the Indian Supreme Court had read down the Aadhaar Act in significant ways (for example, by holding that in case of an authentication failure, an individual was entitled to present an alternative identification), the Jamaican judges still went on to follow, in granular detail, the judgment and reasoning of Chandrachud J.

Justice Batts and the Primacy of Choice

I begin with Batts J.’s judgment, because of its focus on one straightforward and clear principle, which was also at the heart of Chandrachud J.’s dissent: the primacy of individual choice in a constitutional democracy, and how meaningful choice is at the heart of any understanding of human freedom. The “choice”, in this case of course, involves the right of the individual to choose how to identify herself to the State, from among a range of reasonable alternatives.

After summarising the holding of the Aadhaar Majority, Batts J. cited some of the opening words of Chandrachud J.’s dissent, about how technology was reshaping the relationship between citizens of the State. He then went on to cite the core of Chandrachud J.’s reasoning in great detail, because: “his decision is sufficiently important, and so reflective of my own views, that I will outline the details.” (para 337) In particular, Batts J. cited Chandrachud J.’s views on private party usage, the failure of the State to demonstrate that less intrusive means would not work, the merging of data silos, the doctrine of unconstitutional conditions (paragraph 339), and the general presumption of criminality that entailed nationwide biometric collection (paragraph 340). In closing, he noted that:

In words, which I respectfully wish to adopt as my own, the learned Judge [i.e., Chandrachud J.] summarised the overall constitutional failings of the Aadhaar scheme thus … ‘the technology deployed in the Aadhaar scheme reduces different constitutional identities into a single identity of a 12-digit number and infringes the right of an individual to identify herself or himself through a chosen means. Aadhaar is about identification and is an instrument which facilitates a proof of identity. It must not be allowed to obliterate constitutional identity.'” (paragraph 341)

On this basis, Batts J. therefore held that:

The [NIRA] Act proposes to compel persons to divulge information personal to them. It is the right to choose, whether or not to share personal information, which individual liberty in a free and democratic state jealously guards. The mandatory nature of the requirement as well as the breadth of its scope, and the absence of a right to opt out, are not justified or justifiable in a free and democratic society. If it is intended to prevent corruption or fraud, then it is premised on the assumption that all Jamaicans are involved with corruption and fraud. The danger of abuse by the state or its agencies, and the removal of personal choice, outweigh any conceivable benefit to be had by the community or state. (paragraph 349)

This, it will be noted, is squarely applicable to Aadhaar, which has long been mandatory for all practical effects and purposes.

The Chief Justice, Proportionality, and the Surveillance State

The judgment of the Chief Justice was substantially longer (around 250 pages to Batts J.’s 50), and covered more ground. The Chief Justice began with a detailed analysis of the standard of constitutional review under the Jamaican Charter. His discussion is fascinating in its own right, but for our purposes here, the Chief Justice accepted that the relevant standard was that of proportionality, as articulated by the Canadian Supreme Court in Oakes, and of course, the Indian Supreme Court in Puttaswamy. Crucially, however, the Learned Chief Justice noted that:

I am of the view that this approach provides a proper conceptual framework within which to examine NIRA in order to see whether the provisions in that statute meet the standards indicated by the majority in Puttaswamy (September 26, 2018). I must also say that in the application of the standard I prefer the reasoning of Dr Chandrachud J to that of the majority. (paragraph 151)

(For the multiple errors that the Majority made in applying the proportionality standard, see here).

This, the Chief Justice held, was because:

… I am of the view that the strict application of Oakes is the best way to preserve fundamental rights and freedoms. The majority [i.e., in Puttaswamy] appeared to have taken a more relaxed view. The strict Oakes test makes a more granular scrutiny possible by saying that the court must take account of any deleterious effect of the measure being relied on to meet the objective. Thus the greater the severity of the effect the more important the objective must be, furthermore the measure chosen needs to be shown to be the least harmful means of achieving the objective. (paragraph 177)

With these words, the Chief Justice got to the heart of the deficient legal analysis in the Aadhaar Majority. As I have noted previously, the Aadhaar Majority treated the “least restrictive alternative” element of the proportionality standard in an utterly cavalier fashion, refusing to consider relevant facts, and presenting no analysis of the legal burdens involved. For the Chief Justice, on the other hand (and also for Chandrachud J.), it was a critical element of the test, with the burden lying upon the State (especially because relevant information lay with the State). And, in particular:

In the event that the court is of the view that there is a tie then the claimant must prevail for the reason that in constitutional litigation the attitude of the court must be that the right or freedom prevails unless the violation is clearly justified. This approach ensures that the guarantee given by the Charter is maintained. (paragraph 203)

In this context, the Chief Justice went on to find that the State had provided no evidence that mandatory enrolment with criminal sanctions was the least intrusive way of achieving its goal (the Chief Justice distinguished this from Indian Supreme Court proceedings, where the evidence was litigated in some detail). Noting that there was no real evidence about the scale of the existing problems that necessitated this measure, the Chief Justice therefore held that it failed the test of proportionality. (paragraph 228)

After this, the Chief Justice turned to the nature of biometric systems. Here, he began his analysis with the following observation:

For that [i.e., the analysis of biometric systems] I rely on the judgment of Dr Dhananajaya Chandrachud J in Puttaswamy (delivered September 26, 2018). From reading the judgments in this case Dr Chandrachud J, in my respectful view, demonstrated a greater sensitivity to the issues of privacy and freedom that is not as evident in the judgments of the majority or the other judges who delivered concurring judgments. His Lordship had a clear-eyed view of the dangers of a state or anyone having control over one’s personal information and generally I preferred his approach to the issue over that of the other judges. (paragraph 230)

This, of course, is a very polite and respectful way of saying to the Aadhaar Majority, “you just didn’t get it, did you?” And what was it that the Aadhaar Majority didn’t get? Citing paragraphs 120 to 126 of Chandrachud J.’s dissent, the Chief Justice then observed that:

… his Lordship’s major point was that it is one thing to collect biometric data in the context of a criminal investigation and prosecution but quite another to have extensive biometric data collection outside of that context. The reason is that generally there is extensive and detailed provision regarding the collection and use of biometric information in the criminal law context. So far, in the context of general collection of biometric data outside of the criminal law context, it is likely to result in violations of fundamental rights unless there are very strict and rigorous safeguards because once there is a breach of the database the information taken is unlikely to be recovered in full. (paragraph 234)

Even more importantly, the Chief Justice then cited paragraphs 128 to 131 of Chandrachud J.’s judgment, dealing with identification systems, to observe that:

This passage is highlighting the risk of the combined effect of technology with control over data. Unlike the majority in Puttaswamy (September 26, 2018) who seemed to have taken a rather benign view of this aspect of the matter Dr Chandrachud J destroyed the notion that merely because similar or identical information is already in the possession of the state that in and of itself makes taking of such information again legitimate. His Lordship clearly understood the implication of collecting biographical information, combining it with biometric then automating the process with supporting algorithms. Add to that the possibility of profiling. This scenario translates into great power over the lives of persons especially when that data and technology are in the hands of the state and powerful private actors as in Google, Amazon and the like. Of course, with the latter, the engagement is consensual or at the very least the person can opt out after sometime. What NIRA is proposing is control over vast amounts of data, no opt out and linking the data held in different silos by a unique identification number, thereby reducing anonymity even further and increasing the possibility of profiling and generating new information about the data subject. (paragraph 237)

What the Chief Justice understood – and what Chandrachud J. had understood in his dissent – was, of course, the existential peril posed to freedom and privacy that stemmed from merging silos of information, and the ways in which that could be used to generate entirely new sets of information, as well as facilitate profiling. As the Chief Justice observed, devastatingly, “respectfully, the majority in Puttaswamy (September 26, 2018) did not seem to have a full understanding of this and its implications in the say that has been demonstrated by Dr Chandrachud J.” (paragraph 238) On the basis of this bedrock of analysis, then, the Chief Justice went on to hold that NIRA unjustifiably and disproportionately impacted privacy. And in the course of his analysis, he also echoed Batts J., noting – crucially for our purposes – that “the most remarkable thing is that no submission was made to indicate how, for example, a voluntary scheme would prevent the state from providing reliable, safe and secure identification to its citizens or ordinary residents who wish to be part of the scheme.”

Lastly, the Chief Justice also found that Chandrachud J.’s dissenting opinion better articulated the shape and design of a data protection authority that could pass constitutional muster:

I adopt the following paragraph from his Lordship’s judgment at paragraph 236 and apply with such modifications are necessary for application to NIRA. His Lordship stated: An independent and autonomous authority is needed to monitor the compliance of the provisions of any statute, which infringes the privacy of an individual. (paragraph 247(88)).

The absence of any such authority was, therefore, another reason to hold the statute unconstitutional.


The Chief Justice set out his conclusions from paragraphs 245 to 254, striking down substantial portions of the Act, and then holding that they could not be severed – and therefore, the statute as a whole had to fail. Like the judgment of Batts J., the Chief Justice’s judgment is a fascinating study in its own right – in particular, in its assessment of the specific details of the NIRA system, in how it deals with the probabilistic character of biometric identification, the articulations of standards and burdens, the discussion of proportionality and the rebuttal of the State’s arguments that the Court must stay out of policy domains, and – very interestingly – in its disagreement with Bhushan J.’s concurring opinion on Aadhaar, on whether “national security” could be a ground to divulge personal information. Refreshingly, the Chief Justice held that the use of such terms was nothing more than a “Trojan Horse”, which would make the entire point of protecting fundamental rights illusory.

As the above discussion illustrates, however, for us, the most fascinating aspect of this judgment is the in-depth dialogue it undertakes with both the Majority and the Dissent in Aadhaar, on their own terms – and the rigorous and detailed reasons it gives for choosing to follow the Dissent over the Majority.

To end with Hughes again. His full quote reads:

A dissent in a court of last resort is an appeal to the brooding spirit of law, to the intelligence of a future day when a later decision may possibly correct the error into which the dissenting justice believes the court to have been betrayed.


But there are some occasions in history when the betrayal into error is recognised not within, but without. Sometimes, we need friends and colleagues in other parts of the world to hold up the mirror that we are unwilling or unable to look into. Perhaps it is the fate of the Aadhaar Dissent to travel around the world, a light in dark places, long before it is recognised by the brooding spirit of law in its homeland, and the error is corrected at last.

One can but hope.

(Disclaimer: The author was involved on the side of the Petitioners in the Aadhaar challenge.)

Notes from a Foreign Field: The Constitutional Court of Lesotho Strikes Down Criminal Defamation

In a terse, 37-page judgment delivered last week, the Constitutional Court of Lesotho struck down criminal defamation as unconstitutional. The judgment repays some scrutiny, because many of the arguments raised – and addressed – by the Court are similar to the arguments that were made in the unsuccessful constitutional challenge to criminal defamation in India two years ago. Further, the Lesotho Constitutional Court’s application of the global proportionality standard to invalidate criminal defamation shows a potential path forward here, where proportionality has come to the fore after the judgment in Puttaswamy.

The constitutional challenge in Lesotho arose out of criminal proceedings against a satirical article that mocked the Commander of the Lesotho Defence Forces. The author of the article was accordingly prosecuted. Section 104 of the Lesotho Penal Code defined criminal defamation in terms somewhat similar to the IPC. The relevant defences – also akin to the IPC – included proving that the material was true and for public benefit, or establishing legal privilege.

Section 14 of the Constitution of Lesotho guaranteed the right to freedom of speech and expression. The section also authorised restrictions upon the freedom of speech, including, inter alia, “for the purpose of protecting the reputations, rights and freedoms of other persons (interestingly, the Section also provided for a right of reply).

The Court began its analysis by noting the intrinsic and instrumental benefits of the freedom of expression in a democracy (paragraph 8). These are well-worn by now, and do not need repetition. In particular, the Court focused on the importance of satire in a democracy, and the need for any guarantee of the freedom of speech to protect satire (paragraph 9) – especially where “public figures” such as the (former) Commander of the Lesotho Defence Force were concerned, who ought to “display a higher degree of tolerance” (compare this with the Indian Supreme Court’s invented doctrine of “historically respectable personalities). Within this framework, the Court then applied the three-step proportionality standard to decide whether criminal defamation was constitutional. This standard – in the form that the Court endorsed – requires, first, that a restriction upon rights be imposed only through a law, which has a rational connection with the goal; secondly, that the law impair rights only to the minimal extent necessary to achieve the State interest; and thirdly, that there be an overall balancing between the extent to which the right is infringed, and the importance of the goal (paragraph 16).

Criminal defamation cleared the first hurdle, which was a rational connection with the legitimate State interest of protecting reputation. However, it fell at the second hurdle, that of minimal impairment. The Court found, first, that the law was over-broad and vague, inter alia, because the defence of “public benefit” had not been defined, and that “anything could be characterised as not being for “public benefit” due to the elasticity of this concept” (paragraph 18). In particular, by using this concept as a filter, the Court noted that “the Legislature has granted an unfettered discretion to the Prosecutorial authorities”, which would inevitably cast a chilling effect upon freedom of speech and expression (paragraph 18). The Court also found that through its requirement of “truth”, the Section effectively criminalised satire which, by its nature, “exaggerates and distorts reality” (paragraph 18).

Next, the Court held that criminal defamation also failed on the third prong of proportionality – that of a balance between the goal of protecting reputation, and its curtailment of speech. This included the very real possibility of self-censorship, and the existence of civil remedies (paragraph 19), which helped to achieve the same goal without the stigma, direct targeting, and greater punishments that defined the criminal legal regime. The Court closed with noting that the international trend – from a recent judgment of the African Court on Human and Peoples’ Rights (paragraph 21) to international legal instruments (paragraphs 22 – 23) – was towards holding that criminal defamation was no longer consistent with the requirements of democratic societies. The Court therefore concluded:

“The means used to achieve the purpose of protecting reputation interests, in some instances, are overbroad and vague in relation to the freedom of expression guarantee in Section 14 of the Constitution. Furthermore, having concluded that criminal defamation laws have a chilling effect on the freedom of expression, and that, civil remedies for reputational encroachment are more suited towards redressing such reputational harm, I have come to the conclusion that the extent of the above-mentioned sections’ encroachment on the freedom of expression is “not reasonable and demonstrably justified in a free and democratic society.”” (paragraph 24)

Criminal defamation was, accordingly, struck down.

In light of the Lesotho Constitutional Court’s judgment, and before it, in recent times, the judgment of the High Court of Kenya and the African Human Rights Court (both holding criminal defamation to be unconstitutional, in different ways), the Indian Supreme Court’s rambling, near-incoherent, 268-page judgment in Subramanian Swamy v Union of India (2016), which invented new doctrines such as “constitutional fraternity” in order to uphold criminal defamation as constitutional, seems more and more anachronistic. That apart, however, the Lesotho judgment suggests a way forward: in Swamythe two-judge bench of the Supreme Court did not examine the constitutionality of criminal defamation on the three-step proportionality standard. Many of the arguments made before the Lesotho constitutional court – including overbreadth and vagueness, the chilling effect of terms such as “public good”, and the disproportionality of criminal remedies, were dismissed by the Court without a serious examination under the proportionality standard. In Puttaswamy, however, a nine-judge bench of the Supreme Court definitively incorporated the global proportionality standard into Indian constitutional law. In addition to Swamy’s failure to consider may relevant constitutional arguments (see here), this now makes the case for revisiting that judgment even stronger.

In Puttaswamy, the Court acknowledged – within the short span of five years – that its judgment in Koushal v Naz Foundation had been a mistake. This is to the Court’s credit. It would be equally to its credit to acknowledge that its judgment from two years ago, in Subramanian Swamy, was as grave a mistake – and to join the growing ranks of post-colonial countries that have consigned this anachronistic provision to the dustin of history.

Notes from a Foreign Field: The High Court of Kenya Strikes Down Criminal Defamation

In an interesting judgment handed down yesterday, the High Court of Kenya held that criminal defamation unjustifiably restricted the right to freedom of speech and expression, and consequently, was unconstitutional and void. The judgment is part of a growing worldwide trend (with a few noticeable exceptions) to decriminalise defamation, whether judicially or legislatively.

The Constitutional Provisions

Article 33(1) of the Kenyan Constitution guarantees the right to freedom of speech and expression. Article 33(2) provides that this right shall not extend to propaganda for war, incitement to violence, hate speech, or advocacy of hatred. Article 33(3) provides that every person must “respect the rights and reputation of others.”

In addition, Article 24 of the Constitution contains a general limitation clause that states:

“A right or fundamental freedom in the Bill of Rights shall not be limited except by law, and then only to the extent that the limitation is reasonable and justifiable in an open and democratic society based on human dignity, equality and freedom, taking into account all relevant factors, including— (a) the nature of the right or fundamental freedom; (b) the importance of the purpose of the limitation; (c) the nature and extent of the limitation; (d) the need to ensure that the enjoyment of rights and fundamental freedoms by any individual does not prejudice the rights and fundamental freedoms of others; (e) and the relation between the limitation and its purpose and whether there are less restrictive means to achieve the purpose.” 

It is important to note that Article 24 lays down a test of proportionality, which is broadly similar to the test laid down by the Indian Supreme Court in State of Madras vs V.G. Row, where the Supreme Court had held that under Article 19 of the Indian Constitution, the relevant test required consideration of:

“The nature of the right alleged to have been infringed, the underlying purpose of the restrictions imposed, the extent and urgency of the evil sought to be remedied thereby, the disproportion of the imposition, the prevailing conditions at the time, should all enter into the judicial verdict.”

As we can see, there is an overlap between four of the five prongs of the two tests.

The Court’s Analysis

The Petitioners argued that “criminal libel is not a reasonable or justifiable restriction on freedom of expression and added that it is a “disproportionate instrument for protecting the reputations, rights and freedoms of others” and that the remedy in tort is sufficient and less restrictive means of achieving the purpose.” (pg. 3) To buttress this submission, they cited comparative law, including the judgment of the African Court on Human and Peoples’ Rights, the judgment of the High Court of Zimbabwe, and the 2008 Report of the U.N. Special Rapporteur on Freedom of Opinion.

The Court agreed. First, it held that criminal defamation was not saved by Article 24, because the general limitations clause was clearly intended to protect social interests, while criminal defamation was intended to protect an individual interest (the interest of the person defamed). To substantiate this argument, the Court applied the doctrine of noscitur a sociis (a word is known by the company it keeps), on the authority – interestingly – of the Indian Supreme Court. Consequently:

“It is to be borne in mind that defamation of an individual by another individual is a civil wrong or tort, pure and simple for which the common law remedy is an action for damages. It has to be kept in mind that fundamental rights are conferred in the public interest and defamation of any person by another person is unconnected with the fundamental right conferred in the public interest and, therefore, Section 194 out to be construed outside the scope of Article 24 of the Constitution which in my view aims at largely protecting public interest.”

And, with respect to Section 33:

“Section 194, which stipulates defamation of a private person by another individual, has no nexus with the fundamental rights conferred under article 33 of the Constitution, for Article 33 is meant to include the public interest and not that of an individual and, therefore, the said constitutional provision cannot be the source of criminal defamation. I base this argument on two grounds:- (i) the common thread that runs through the various grounds engrafted under Article 33 (2) (a)-(d) are relatable to the protection of the interest of the State and the public in general and the word “defamation” has to be understood in the said context, and (ii) the principle of noscitur a sociis, when applied, “defamation” remotely cannot assume the character of public interest or interest of the crime inasmuch a crime remotely has nothing to do with the same.”

(There is a parallel worth thinking about here, because the Indian Supreme Court has often held that the purpose of Article 19(2) is to protect “social interests“.)

However, given that Section 24 spoke about the “rights of others“, and Section 33 spoke about “reputation”, was that not a basis for the constitutionality of criminal defamation?  The Court responded that the question was whether criminalising defamation was a proportionate method of protecting the rights of others. It held that it was not. To start with, the Court observed:

“Human rights enjoy a prima facie, presumptive inviolability, and will often ‘trump’ other public goods.” (p. 8)

Within this framework, the Court held that the question of proportionality would have to be answered in two phases: “firstly, what are the consequences of criminalizing defamation and, secondly, is there an appropriate and satisfactory alternative remedy to deal with the mischief of defamation.” (p. 11)

On the first issue, the Court focused on the specific aspects of the criminal process: “The practical consequences that would ordinarily flow from a complaint of criminal defamation are as follows; the accused person would be investigated and face the danger of arrest. This would arise even where the alleged defamation is not serious and where the accused has an available defence to the charge. Thereafter, if the charge is prosecuted, he will be subjected to the rigors and ordeal of a criminal trial. Even if the accused is eventually acquitted, he may well have undergone the traumatizing gamut of arrest, detention, remand and trial. Moreover, assuming that the accused has employed the services of a lawyer, he will also have incurred a sizeable bill of costs which will normally not be recoverable.” (p. 11)

While, admittedly, this problems would afflict any person accused of any criminal offence, the case of free speech was crucially different because of the chilling effect. According to the Court:

“The overhanging effect of the offence of criminal defamation is to stifle and silence the free flow of information in the public domain. This, in turn, may result in the citizenry remaining uninformed about matters of public significance and the unquestioned and unchecked continuation of unconscionable malpractices.” (p. 11)


“The chilling effect of criminalizing defamation is further exacerbated by the maximum punishment of two years imprisonment imposable for any contravention of section 194 of impugned section. This penalty, in my view, is clearly excessive and patently disproportionate for the purpose of suppressing objectionable or opprobrious statements.” (p. 11)

Furthermore, if proportionality was about ensuring that the least restrictive method was applied to serve a particular goal, then the very existence of an equivalent civil remedy made criminalising the offence disproportionate. The Court held:

“I am clear in my mind that there is an appropriate and satisfactory alternative civil remedy that is available to combat the mischief of defamation. Put differently, the offence of criminal defamation constitutes a disproportionate instrument for achieving the intended objective of protecting the reputations, rights and freedoms of other persons. Thus, it is absolutely unnecessary to criminalize defamatory statements. Consequently, I am satisfied that criminal defamation is not reasonably justifiable in a democratic society within the contemplation of article 24 of the Constitution. In my view, it is inconsistent with the freedom of expression guaranteed by 33 of that Constitution.” (p. 14)

Finding that this view was also in accord with international practice as well as the decisions of the African Court, the Kenyan High Court struck down criminal defamation.

Comparisons with India

It is interesting to note that all the arguments that proved decisive with the High Court of Kenya, were argued before the Supreme Court last year in Subramanian Swamy vs Union of India – and almost completely ignored. It was argued that defamation was primarily aimed at protecting individual reputation, and therefore inconsistent with the very purpose of criminal law (to provide public remedies). It was argued that criminalising defamation was a disproportionate response under Article 19(2), because of the nature of the criminal process. And it was also argued that the Court was required to be particularly solicitous to the question of balance as far as the freedom of speech was concerned, because of the reality of the chilling effect. However, instead of engaging with these issues, the Court decided to elevate reputation to the status of the right to life, invented a doctrine of “constitutional fraternity” out of thin air, and upheld criminal defamation in a rambling, 270-page long judgment, which was notable for its failure to address the precise arguments that – as pointed out above – the High Court of Kenya found convincing.

It is not only for its verdict, but also for its lack of reasoning, that Subramanian Swamy needs an urgent rethink. The Kenyan High Court’s terse and lucid 14-page judgment provides us with a good template of what such a rethinking might look like.


Notes from a Foreign Field: The New Zealand High Court Issues Its First “Declaration of Incompatibility”

(We are starting a new series called ‘Notes from a Foreign Field’, focusing on decisions of other constitutional courts, and constitutional controversies in other jurisdictions, written by specialists from those jurisdictions. In the opening post, Max Harris, a New Zealand lawyer and presently Prize Fellow at All Souls College, Oxford, writes about a recent, important decision of the New Zealand High Court, which broke new ground in the area of judicial review)

On 24 July 2015, Justice Heath of the New Zealand High Court issued a landmark human rights decision, Taylor v Attorney-General [2015] NZHC 1706. The case is worth reviewing for readers outside of New Zealand. it provides an overview of the human rights landscape of a jurisdiction that is often overlooked, presents a further perspective for global debates on prisoner voting, and is an example of robust judicial reasoning in a constitutional context.

The Taylor case arose out of New Zealand’s Electoral (Disqualification of Sentenced Prisoners) Amendment Act 2010, which imposed a blanket ban on prisoners voting in New Zealand elections. The position prior to 2010 had been that prisoners serving a prison term longer than three years were banned from voting. Arthur Taylor, a prisoner, challenged the 2010 Act. He argued that it posed an unreasonable limit on his right to vote, under s 12 of the New Zealand Bill of Rights Act 1990.

The New Zealand Bill of Rights Act 1990 is an early example of a statutory bill of rights. The original draft Bill of Rights empowered judges to strike down legislation, but after public opposition to this, the Act reached a compromise solution. It lists a standard set of rights and freedoms, indicates that rights are subject only to “reasonable limits” that can be “prescribed by law” and “demonstrably justified in a free and democratic society” (s 5), but makes clear that legislation cannot be struck down where limits on rights are found to be unreasonable (s 4). The Act allows the Attorney-General to flag up violations of rights at the legislative drafting stage (s 7), and also notes that “[w]herever” legislation “can be given a meaning” consistent with rights, “that meaning shall be preferred to any other meaning” (s 6).

The Act was a model for the UK’s Human Rights Act 1998. However, unlike the Human Rights Act, the New Zealand Bill of Rights Act 1990 does not spell out the remedies available for litigants in the event that a court finds that legislation unreasonably limits rights. In Simpson v Attorney-General [1994] 3 NZLR 667 (Baigent’s case), the New Zealand Court of Appeal found that damages should be available for violations of the New Zealand Bill of Rights Act. What remained in doubt was whether New Zealand courts could issue a declaration of inconsistency or incompatibility (of the kind explicitly allowed by s 4 of the Human Rights Act 1998 in the UK), where an Act imposes an unreasonable limit on rights. That was the key issue in the Taylor case, because Arthur Taylor asked the High Court to issue a declaration of inconsistency with respect to New Zealand’s prisoner disenfranchisement legislation.

Let us consider the judgment. Justice Heath points to the fact that interestingly, the Crown had conceded that there was inconsistency between the legislation and the right to vote, and agrees with the Attorney-General’s preliminary opinion that there was an inconsistency. He adds one further reason why the legislation is an unreasonable limit on the right to vote: it arbitrarily focuses on imprisonment, rather than conviction, thereby allowing a person who is sentenced to home detention to retain a right to vote, though that person may be as equally culpable as another person sentenced to imprisonment.

So far, so uncontroversial. The real question in the case, however, was that given the acknowledged inconsistency, whether the Court has jurisdiction to grant a declaration of inconsistency, in light of the New Zealand Bill of Rights Act 1990 and relevant policy concerns.

The first argument made by Crown lawyers was that a declaration could not be issued in a case where there is no dispute over interpretation of legislation. It was said that the main remedy provided by the New Zealand Bill of Rights Act is an interpretive one: courts can try to interpret away a possible inconsistency with rights, but cannot issue a declaration saying that legislation is inconsistent with rights. Justice Heath considers this argument and rejects it. He accepts that there are some restrictions on when a declaration can be granted. The New Zealand District Court is a creature of statute and cannot grant any declarations (let alone a declaration of inconsistency), and declaratory relief should not be available in a criminal trial, because a declaration represents civil relief that would be inappropriate in a criminal context. (He cites a Court of Appeal decision that notes the inappropriateness of using civil remedies in a criminal context.) But he suggests that these should be the only restrictions placed, in principle, on the issuing of declarations of inconsistency.

Justice Heath points out that earlier courts had said that judges can, and indeed sometimes must, indicate an inconsistency between legislation and the New Zealand Bill of Rights Act. He says further that to allow a declaration of inconsistency would not contradict s 4 of the New Zealand Bill of Rights Act 1990 – the provision preventing judges from striking down legislation on Bill of Rights grounds. Acknowledging the room for “judicial choice”, Heath J reviews earlier case law where remedies (including damages) for Bill of Rights breaches have been developed. He extracts a general principle that “where there has been a breach of the Bill of Rights there is a need for a Court to fashion public law remedies to respond to the wrong inherent in any breach of a fundamental right”. He concludes that Parliament did not intend to exclude the ability of a court to make a declaration of inconsistency.

Justice Heath feels fortified in this conclusion by the fact that a legislative amendment in 2001 allowed declarations of inconsistency in discrimination cases (heard by the Human Rights Review Tribunal in New Zealand). Through this legislative act, Justice Heath says, “Parliament has signaled that it sees no particular objection to that particular remedy being granted”. It would be odd for Parliament to confer this power on a lower tribunal, notes Justice Heath, and to empower higher courts to review use of this power on appeal, but to remove the right of higher courts to issue declarations of inconsistency. Whether a declaration of inconsistency breaches art 9 of the 1688 Bill of Rights (which protects parliamentary privilege and remains part of New Zealand law) or principles of comity between the legislature and the courts are matters that only affect whether a declaration should be issued in a particular case, according to Heath J, not matters that go to the general jurisdiction of a court to issue a declaration.

Addressing whether a declaration of inconsistency is appropriate in the Taylor case, Heath J considers arguments based in the Bill of Rights 1688 and comity. Heath J states that if courts are able to give reasons why legislation imposes unreasonable limits on rights under s 5 of the New Zealand Bill of Rights Act, it is hard to see how a declaration would create any greater intrusion on parliamentary privilege or comity. This is a kind of boot-strapping argument: if s 5 of the New Zealand Bill of Rights Act is constitutionally legitimate (a proposition Heath J assumes), then a declaration of inconsistency must be similarly legitimate.

Courts should not hold back from issuing declarations out of fear that they might be ignored, says Heath J. The New Zealand judicial oath requires decision-making “without fear or favour”. And the Court is, after all, not seeking to persuade – it is merely stating the law (echoing Justice Marshall’s statement from the US Supreme Court decision in Marbury v Madison (1803) 5 US 137). Heath J disagrees with the comments of an earlier judge in an interlocutory decision in the Taylor case that a court might hesitate to issue a declaration where the Attorney-General has already flagged up a Bill of Rights inconsistency in a s 7 report to Parliament. There is no reason why a court “should not reinforce the Attorney’s report”, notes Heath J. He adds that a court should also be able to disagree with an Attorney-General’s report.

Should the absence of a live controversy between parties prevent a declaration? Heath J points out that there is no limit of this kind for ordinary declarations under the Declaratory Judgments Act 1908 (though he doubts whether a declaration of inconsistency could be granted under that Act). Points of “constitutional importance” should be ventilated, says Heath J: “[t]he importance of the right and the nature of the inconsistency are sufficiently fundamental to demand a remedy”.

In this case, Heath J confirms that a declaration of inconsistency will be granted. The case concerns a central aspect of democracy, the right to vote: “if a declaration were not made in this case, it is difficult to conceive of one in which it would”. Heath J notes that “a formal declaration” is more appropriate than “an observation buried in [a court’s] reasons for judgment”. There is no violation of Art 9 of the Bill of Rights 1688 or principles of comity; the comment is on “the consequences of a legislative act”, not the internal workings of Parliament itself. The functions of the Attorney-General’s pre-legislative report and the court are different; the Attorney-General is considering an apparent inconsistency, a court is considering an actual inconsistency. A court’s ruling will also be more accessible. Finally, Heath J says, Parliament’s earlier legislative recognition of declarations in discrimination cases shows a certain amount of approval for the notion of declarations of inconsistency. Heath J notes in passing that there are “powerful arguments” that the earlier limitation on prisoner voting (allowing the vote only for prisoners serving fewer than three years in prison) could be Bill of Rights-compliant. He concludes with the declaration itself, in the following terms:

Section 80(1)(d) of the Electoral Act 1993 (as amended by the Electoral (Disqualification of Sentenced Prisoners) Amendment Act 2010) is inconsistent with the right to vote affirmed and guaranteed in s 12(a) of the New Zealand Bill of Rights Act 1990, and cannot be justified under s 5 of that Act.

This is the first case in New Zealand in which a judge has issued a declaration of inconsistency. (In an earlier Court of Appeal case, R v Poumako [2000] 2 NZLR 695 (CA), one judge, Thomas J, issued a declaration of inconsistency, and reviewed the arguments for declarations in some detail. But he was the sole dissenting judge in this case.) The declaration has not resulted in any legislative reconsideration of the prisoner voting ban, however. It seems that no law change will be forthcoming. The Crown has not appealed the ruling.

What more general points, then, can be drawn from this detailed review of Heath J’s reasoning in Taylor v Attorney-General?

First, there are parts of the judgment at which criticism might be directed. Given Heath J’s emphasis on how distinct the Attorney-General’s s 7 report is from a court’s later review of legislation, it is surprising that he does not undertake a fresh proportionality assessment of the prisoner voting legislation in this case. Perhaps Heath J felt that in a controversial case like this one, and as the first judge ever to issue a declaration of inconsistency, it would be safer simply to affirm the Attorney-General’s earlier reasoning.   However, it would have been helpful for Heath J to offer further reasoning on this point, especially since prisoner voting bans have been contentious in other jurisdictions, including the United Kingdom, South Africa, and Canada. (An earlier interlocutory decision of Brown J did refer to some of these other cases.) As well, Heath J is a little peremptory in some conclusions. He is quick to accept that declarations of inconsistency should not be issued in criminal trials, when there is no legislative reference to this carve-out. And he is not entirely convincing in his claim that declarations of inconsistency do not undermine Art 9 of the Bill of Rights. Heath J might also have made some broader comments about the proper approach to the separation of powers and dialogue under the New Zealand Bill of Rights Act. Perhaps, however, these are points that might be expected in the judgment of an appellate court. Overall, Heath J’s judgment is admirably careful, considered, and courageous.

Secondly, the effect of the judgment – even if it is not momentous, and only slightly shifts the pre-existing position – is to recalibrate subtly the relationship between the courts and Parliament in New Zealand. New Zealand is a jurisdiction where judges have restricted powers. There is no single codified constitution in New Zealand, just as there is no single codified constitution in the United Kingdom. The generally accepted position is that judges cannot strike down legislation, and parliamentary sovereignty is often invoked. Judges (with some notable exceptions) tend to be deferential towards the executive and the legislature. Against that backdrop, this judgment gives judges slightly greater powers in human rights cases and should cause Parliament to hesitate a little more when passing legislation that might violate human rights. Whether, of course, Parliament actually shows more respect for human rights as a result of this judgment is an empirical question. The early signs are not especially promising: New Zealand Justice Minister Amy Adams, after the judgment was released, said that she was considering the judgment, but there seems to have been no further comment from the Minister since July of this year. Opposition Labour and Green Parties did use the judgment to call for the prisoner voting legislation to be repealed, and this highlights a further benefit of declarations of inconsistency: even if they do not lead to direct political change, they can provide tools for citizens, campaigning groups, and other politicians to criticise legislation.

In the earthquake-prone islands of New Zealand in the South Pacific, this judgment may not have shifted the tectonic plates of constitutional law – but at the very least, Taylor has jolted the constitutional landscape. The case is a significant milestone in the development of the jurisprudence of the New Zealand Bill of Rights Act 1990, and an important reminder of the valuable role that courts can play in clarifying matters of principle – and upholding human rights.