top of page

How Not to Do Comparative Constitutionalism: Lessons from South Africa

Kushagr Bakshi *[1]

While there has been an interest in comparative constitutional research for quite some time in academia and in the courts, methodologies of comparative work remain a big source of debate. While some suggest contextual analysis, others tend to prefer empirical analysis as a method of engaging in comparative constitutional research.[2] However, courts have tended towards comparative judicial reasoning, importing constitutional standards of review from other jurisdictions.[3] South Africa, being a curious blend of an Anglo-Saxon vision of common law (adjudicated on an empirical case-by-case basis) and the system of Roman-Dutch law (deductive continental analysis) has been engaging in comparativism since well before the new Constitution came into effect. In S v. Makwanyane, Justice Chaskalson set out an emphatic defence of the comparative tradition, arguing that courts ought not to treat foreign decisions as precedent, rather that they use the lessons and reasoning (contextually considered) from foreign courts as illuminating texts to guide their own reasoning. In what follows, I argue that the Constitutional Court of South Africa in Rafoneke & Others. v. Minister of Justice and Correctional Services and Others, failed to heed this call, instead using decisions by the Canadian and Indian Court as blind precedents with limited focus on the reasoning, context and nature of the rights in question. In addition to arriving at a full criticism of the court’s reasoning, I also hope to lay out an example on how not to do comparative constitutionalism.

Foreign Lawyers and Registration

Briefly, the case before the South African Court concerned Section 24(2) of the Legal Practice Act, which permitted only citizens and permanent residents to register as legal practitioners in South Africa. The appellants, individuals from Lesotho and Zimbabwe, who earned their law degrees and completed their articles of clerkship in South Africa but did not hold permanent residency, argued that this provision unfairly discriminated against them and violated their right to equality as guaranteed under Section 9 of the South African Constitution. The Court, using the test laid down in Harksen v. Lane, engaged in a rational basis review and held that protectionism was a legitimate government aim and foreigners who did not have permanent resident status did not have the same fixity of connection to the country, thereby stating that the differentiation did not amount to discrimination. They went on to the next stage of review, holding that while differentiation based on citizenship could affect fundamental human dignity, the current regulation did not prevent the appellants from seeking alternative sources of employment, and hence they were not left destitute (and have suffered no hardship: a factually incorrect conclusion), thereby concluding that the differentiation did not amount to unfair discrimination.

Within the South African Constitutional structure itself, this judgement is extremely questionable. For example, in Prinsloo, the court had held that any legitimate government purpose must be a defensible vision of the public good and that the burden was on the government to justify such differentiating measures. However, in this case, the court was content to find that the government has no duty to extend to foreigners the right to trade or occupation and refused to engage in a substantive review of the governments ‘legitimate purpose’, as Prinsloo warrants. Any further enquiry would have led the court to engage with the Minister’s patently ridiculous submissions that this provision was aimed at preventing individuals from circumventing immigration laws (as a grant of a law license does not change immigration status, it’s hard to see the ‘rational nexus’) and that foreign lawyers were untrustworthy.[4] Further, the Harksen test in its last stage, requires the court to consider ‘unfairness’ as demonstrated by impact on the aggrieved party. Instead of engaging in proportionality analysis to consider the effectiveness of such a measure, whether it was the least restrictive recourse and the disproportionate harm it caused to the appellants, the court was content with saying that since they had access to some jobs and were not destitute, there was no unfairness. While many South African lawyers have criticised the court’s judgement and its reasoning, I intend to focus on the court’s usage of foreign law cases, namely Skapinker and Balaji within its rational basis analysis.

Poor Comparativism

At the outset, it is perhaps relevant to point out that despite my criticism, the South African Court hasn’t strayed too far from the grain here. Many jurisdictions around the world have similar restrictions and do not permit individuals, who are not citizens or permanent residents, to register with the relevant licensing authorities in legal practice and other professional contexts. Picking up on this trend, the Court tried to rely on a couple of cases from around the world, to augment its reasoning. However, it made two crucial errors in doing so. First, it chose poor cases with limited direct applicability, and second, instead of trying to incorporate the reasoning of the courts in these judgements, it simply cited them for evidence of a State’s regulatory authority, thereby evading the legal burden incumbent upon them.

On Poor Choices

The first case the court cites is a Canadian case which involved a foreign citizen bringing the first case under the Canadian Charter of Rights and Freedoms to their Supreme Court. Joel Skapinker, (ironically a South African citizen) who had applied to the Ontario bar for registration, was denied admission as the relevant act required that he be a Canadian citizen. Skapinker challenged the constitutional validity of this provision, but unlike Rafoneke, brought his challenge under Section 6(2)(b) of the Charter, alleging that his right to mobility included within it, a right to work. In rejecting his claim, the Canadian Supreme Court focused on how Section 6(2)(b) did not avail Skapinker of an independent constitutional right to work. But the appellants in Rafoneke were not claiming any independent right to work, rather an equality violation, which requires a substantially higher burden of review given the transformative character of the South African Constitution. The South African Constitutional Court did not consider this crucial difference when citing Skapinker. This error becomes all the more apparent when one looks at Andrews, where the Canadian Supreme Court struck down a similar citizenship requirement, but under Section 15 (the equality provisions) of the Charter, holding that a citizenship requirement was indeed differential treatment and unjust. This is a particularly egregious error because the South African Court cites Andrews later in its judgement when holding that non-citizens lack political power and therefore differentiation based on nationality can impair dignity.

The judgement in Balaji too has little relevance to the situation in Rafoneke. In Balaji, the Indian Supreme Court dealt with the issue of whether foreign law firms were allowed to practice in India on a visiting or consulting basis without registering with the Bar Council. The situation exclusively dealt with foreign lawyers who did not register under the Advocates Act, while pursuing consultancy and arbitration work. Importantly, there was no rights based constitutional challenge in this case and the Supreme Court’s decision was premised on the principle of reciprocity, holding that foreign lawyers could not practice in India without their home countries observing this principle. In Rafoneke, this was already established, as foreign lawyers registered as legal practitioners in their home jurisdictions were permitted to retain their rights and continue practicing in South Africa. While the appellants did challenge the differentiation between themselves and such foreigners, the Court referring to the obligations under General Agreement in Trade and Services and reciprocity, drew a distinction between the rights of those already admitted to practice and the rights being claimed by the appellants, using an intelligible differentia test.

On Poor Usage

But if, as I have argued, these cases bear limited relevance to the case at hand, then how did the Court use them? The Court presented these cases at the end of the section on rational basis review. Through paragraphs 87 to 89, the Court briefly laid out the fact pattern and results of each of these cases and then at the end of paragraph 89, argued:

Both these cases fortify the reasoning that law may be enacted to regulate entry into a profession and States are entitled to restrict such entry on the basis of citizenship.

On simple analysis, it appears that the Court is using these examples to show how many jurisdictions restrict the practice of the law to citizens and hence, it is legitimate for South Africa to do the same. This would be a poor example of comparativism and exactly what Justice Chaskalson warned against doing. For, the argument for comparativism is not to blindly follow what others do, but to look at the reasoning of various courts as instructive manuals on how they dealt with similar problems. But I do not believe that is what the Court is doing. Instead, the Court is using these cases to show that prima facie, within their sovereign authority, states have the power to regulate entry into a profession, specifically the practice of the law and that the court need not (cannot?) evaluate the legitimacy of the underlying reasons for such restrictions.

Here’s the kicker. Nowhere in this case have any of the appellants or amici argued that the State does not, prima facie, have the authority to regulate entry into a profession. In fact, while Section 22 of the Constitution of South Africa guarantees every citizen the right to trade, practice and profession, it subjects that right to the power of regulation by law. It seems that this was painfully obvious to the Court as well, for in paragraph 74, the Court states that it did not understand the complaint to be that a sovereign state has no power to pass laws regulating a profession or trade. Instead, the argument from the appellants has always been that the differentiation between them and permanent residents and citizens does not have a rational basis and does not serve any legitimate government purpose. Instead, what the court does is that in paragraph 78 it argues that the state has a right to admit foreigners on conditions the state deems fit to prescribe and that no country has the duty to extend freedom of trade to non-citizens. Since it has a right to restrict trade and no duty, the court holds this restriction to be legitimate, without an underlying review into the purpose, thereby not considering any of the government’s justifications.[5] In effect, the use of comparative examples here, is a dodge by the Court through which it refuses to engage with the substantive arguments of the appellants regarding whether this discrimination serves a legitimate government purpose.

What lessons can we draw from the court’s reasoning in this case? First, the choice of which cases the courts and researchers refer to, is just as important as the cases they chose to exclude from their frame of comparative reference.[6] Second, using the results of cases across jurisdictions, as support for your argument or decision, without paying attention to the context or judicial reasoning, adds little validity to your own argument or decision. Finally, it is important to situate the comparative examples within the legal test or reasoning, otherwise you run the risk of obscuring the constitutional test and presuming the validity of the reasoning due to the validity of the result. Essentially, one must guard against the logical fallacy[7] of assuming that because two people arrived at the same result, they must have followed equally sound judicial reasoning. In this sense, the decision of the South African Constitutional Court shows us how not to do comparative constitutional law.


The views expressed above are the author's alone and do not represent the beliefs of Pith & Substance: The CCAL Blog.

[1] The author would like to extend his gratitude to Prof. Karthigasen Govender for his helpful overview of the South African Constitutional Framework, and also to Deputy Chief Justice Maya of the Constitutional Court of South Africa for her discussion on this case. [2] On variance in comparative methodologies, see, Vicki Jackson, Comparative Constitutional Law: Methodologies in The Oxford Handbook of Comparative Constitutional Law 54 (Michel Rosenfeld & Andras Sajo eds., Oxford University Press, 1st ed., 2012). See also, Ran Hirschl, Comparative Methodologies in The Cambridge Companion to Comparative Constitutional Law 11 (Roger Mastermann & Robert Schutze eds., Cambridge University Press, 1st ed., 2019). [3] A prime example of this would be the adoption of proportionality analysis in rights review by courts across the world. See, Moshe Cohen-Eliya & Iddo Porat, Proportionality and Constitutional Culture 24, 43 (Cambridge University Press, 1st ed., 2013). [4] In fairness to the Court, they did not engage at all with this clearly xenophobic argument, considering it faulty. [5] Which as I have argued previously, is the required standard under South African Law. See Prinsloo v. Van der Linde [1997] ZACC 5. [6] In fairness, I doubt this is a novel insight. It was the first thing my supervisor told me when I started comparative research. [7] Akin perhaps to ‘post ergo propter hoc’.


bottom of page