Dual citizen parliamentarians are banned under Australia’s Constitution – even if they don’t know of their foreign citizenship. The far-reaching implications of this are evident from Re Canavan and Re Gallagher.

The dual citizen ban results from the last part of s 44(i) (‘the second limb’), which provides that a person who is “a subject or a citizen or entitled to the rights or privileges of a subject or a citizen of a foreign power” is incapable of being chosen or sitting as a parliamentarian. This overlaps with the first part of s 44(i) banning a person “under any acknowledgment of allegiance, obedience, or adherence to a foreign power” (‘the first limb’).

I argue below that the framers of the Constitution never intended the second limb to apply to dual citizens, as the common law then generally ignored competing foreign claims to the allegiance of the Queen’s subjects.

Why are dual citizen parliamentarians banned?

The High Court’s answer in Re Canavan was that the ordinary meaning of the words requires the ban, and s 44 (i)’s purpose and drafting history does not support any other approach.

But why would the ban have been thought a good idea in 1900, when the Constitution was enacted? No doubt, s 44(i) seeks to protect against parliamentarians who may unduly favour, or be influenced by, foreign powers. It treats foreign citizenship as indicating predisposition to that. Reasonable as that may seem, it’s less so where foreign law makes a person a citizen because a grandparent was a citizen.

All recent decisions of the High Court have accepted as correct the approach taken in 1992 in Sykes v Cleary to the recognition of foreign law regarding nationality. On that approach (’90’s approach’), if an Australian citizen is also claimed as a citizen by foreign law, our law recognises both nationalities and treats the person as a foreign citizen for the purposes of s 44(i) (subject to very limited exceptions). The High Court was not asked to consider the position when the Constitution was enacted in 1900, nor was that obviously relevant. The court applied the common law as it was in 1992, referring to mid-twentieth century cases, and the 1930 Hague Convention, which required recognition of foreign citizenship and dual nationality.

The law regarding nationality in 1900

In 1900 the law regarding nationality (Boll 2007; Spiro 2017) and “conflicts of laws” (Dicey 1896, 15-22, Ch III; Bartholomew 1959) was very different, leading to a different approach to dual nationality (‘1900 approach’). The drafting history of s 44(i), viewed in the light of the 1900 approach, suggests that the second limb of s 44(i) was never intended to apply to “subjects” who were also dual citizens. On that reading, dual citizens would still be disqualified under the first limb if, for example, they took a foreign oath of allegiance or were voluntarily naturalised in a foreign state (Piggott, 1907, 158-9). But the first limb would be unlikely to catch a dual citizen who had done nothing to seek or acknowledge their foreign citizenship.

It is difficult to comprehend the 1900 approach without some sense of the historical context. (References below to British subjects include all we now call Australian citizens, since before 1948 there were no Australian citizens.)

English law’s use of “subject” rather than “citizen” reflected the notion of “subjection to one lord and king” (Salmond (1902) 18 LQ Rev 49). Until 1870, all British subjects owed absolute and indelible allegiance to their sovereign, for life.

For centuries, English common law had based nationality on place of birth, which clashed with the prevailing European approach based on descent. Conflict between these approaches made competing claims to allegiance, and dual nationality, inevitable. Britain, like most states, regarded its claims to a subject’s allegiance as paramount. Nineteenth century empire building ensured that Britain was effectively competing with other countries for much needed subjects/citizens.

These circumstances, counter-intuitively, created a strong incentive for Britain to ignore, or treat as irrelevant, any foreign nationality of its subjects. Britain could not stop foreign states conferring citizenship, and penalising dual citizens themselves would only risk driving away subjects whose allegiance Britain wanted to retain.

The 1812 naval war between Britain and the US illustrates the importance of this issue to Britain. One cause of the war was the British practice of stopping US ships to “impress” naturalised Americans into naval service – ignoring their US citizenship and treating them solely as British subjects owing indelible allegiance to the King. The Prince Regent (later, King George IV) asserted, in response to the US declaration of war, that abandoning this claim would “expose to danger the very foundation of our maritime strength”. The issue was not conceded by Britain in the treaty ending the war, and continued to be a source of tension with the US for at least half a century (1869 Royal Commission report, Appendix 29-52).

Finally, after a lengthy and comprehensive 1869 Royal Commission report, the Naturalization Act of 1870 (Imp) (‘1870 Act’) was passed to permit British subjects to be naturalised in a foreign country and cease to be British subjects. As a result, allegiance was no longer “indelible” for those who left the British empire, but for others, it continued to be absolute and paramount.

Section 7 of the 1870 Act also removed old restrictions on naturalised British subjects (who would likely be dual citizens) becoming members of either house of the Imperial Parliament in London. That was an extension of reforms 25 years earlier which had recognised the importance of such privileges for Britain’s interest in attracting immigrants able to make significant contributions to the nation. This approach was affirmed again in Britain in 1901, following a review underway when the Australian Constitution was enacted.

In 1900, English law generally ignored any foreign nationality of a British subject, or treated it as irrelevant, unless a statute clearly required otherwise.

Peter Nygh – who later authored a seminal Australian text on conflict of laws – commented in 1963:

… it is difficult to imagine any circumstances in which the courts would give effect to the existence of the second nationality except of course where they are expressly directed to do so ….

Clive Parry expressed similar views (Parry 1958, 343, 360, 362, 365). The reason for that approach was explained by W E Hall (Hall 1894, 20):

English law knows no distinction between different classes of natural born British subjects in respect of rights or obligations. All alike, whatever their parentage, have the same duty of allegiance, the same rights within the British dominions, and, subject only to a qualification introduced in certain cases for reasons of public policy, the same right to recognition and protection abroad. [emphasis added]

The statement above relates to natural-born British subjects, but the approach to naturalised British subjects was generally the same when they remained within the British empire. Other leading scholars at the time expressed similar views (Dicey 1896, 174; Westlake 1904, 221-9; Westlake 1905, 356).

The 1900 approach of disregarding foreign nationality of British subjects continued into the early part of the twentieth century at least (Nygh 1963; Parry 1958). Examples include R v Lynch [1903] 1 KB 444 and Ex Parte Freyberger [1917] 2 KB 129. Courts would, of course, still recognise the foreign nationality of a British subject if clearly required to by statute. An example is Kramer v Attorney-General [1923] AC 528 where a British Order expressly charged property of “German nationals” to give effect to the treaty ending World War I. The House of Lords considered that only German law could give meaning to “German national” (at 537), but emphasised that the decision turned on construction of the Order (at 532, 539).

Under the 1900 approach, a British dual citizen would be treated the same as any other British subject unless s 44(i) clearly required otherwise. In contrast to the term “German national” in Kramer, the phrase “subject … of a foreign power” could only be interpreted in the first instance under the law of Australia – since Australia itself would be “foreign” under the laws of any other country. On the 1900 approach, the common law would ignore any foreign nationality of British subjects and treat them only as subjects of the Queen, and therefore, not “foreign”. There would be no need to consider the law of any particular foreign jurisdiction since, under Australian law (as in England), any nationality conferred would be irrelevant in the case of a British subject. To paraphrase Younger LJ (dissenting) in Kramer [1922] 2 Ch D 850 at 877 – while Australian (or English) law could not determine when a person was a subject of a particular foreign power, it could, for its own purposes, determine when that person was not a “subject … of a foreign power”. The House of Lords did not disagree, even though it rejected Younger LJ’s conclusions.

The effect of the second limb of s 44(i) under the 1900 approach would have been limited. However, it would not have been entirely superfluous even though, in 1900, s 34(ii) separately required all parliamentarians to be subjects of the Queen. Arguably, the second limb was needed to ensure that parliamentarians who became foreign citizens after being elected would lose their places under s 45. More importantly, the requirement to be a subject in s 34(ii) only applied until Parliament provided otherwise. Consequently, the second limb was needed to limit how far Parliament could go, without a referendum, if it dispensed with s 34(ii).

The drafting of the Australian Constitution

The drafting of the Constitution was a long and tortuous process taking nearly a decade. Drafts were prepared by a drafting committee and debated by delegates from the federating colonies at four lengthy Convention sessions – the first in 1891 and the rest, after a long break, in 1897-8. The agreed draft was then twice put to “referendum” votes by federating colonies. Finally, the Imperial Parliament in London passed it, with some amendments, in 1900.

The drafting committee did not lack for legal expertise. It was chaired by Sir Samuel Griffith QC in 1891, and by Edmund Barton QC, then Leader of the Convention, in 1897-8. Both chairs drove the drafting and were heavily involved. Griffith, Barton and Richard O’Connor QC, another member of 1897-8 committee, would later comprise the new High Court for its first three years.

The first draft of what became s 44(i) was similar to clauses in the federating colonies’ constitutions. Two amendments to it are worth noting – one made at the start of the process, and one at the very end.

The first changed a provision causing elected parliamentarians to lose their places on doing certain acts, into a provision that made anyone who did those acts “incapable” of being elected (1891 draft s 46). That meant the relevant acts – such as swearing a foreign oath of allegiance – disqualified them indefinitely. A question at the Convention (15 April 1897, 736), and a confidential criticism by the British Colonial Office (Williams 2005, 713, 727), both suggested that the disqualification should be lifted for someone who later became a naturalised British subject.

The second noteworthy amendment was a redraft that created the dual citizen ban. It did that by changing the basis for disqualification from an act to a status – from disqualification after doing a specified act, to disqualification while being a foreign citizen or equivalent. None of the drafts before that amendment, and none of the precedents in the colonies’ constitutions, would have disqualified any of the parliamentarians found to be dual citizens in recent years.

The change from action to status addressed the British Colonial Office’s criticism since it meant that disqualification would end once a person stopped being a foreign citizen. Barton noted on his copy of the British Colonial Office’s comments that its criticism had been addressed. That may have been the reason for the redraft. (Re Canavan discusses the Colonial Office’s criticism (at [33]), but not Barton’s note.)

The other effect of that change was to create a broad dual citizen ban that disqualified British subjects who had done nothing to seek – and might not even have been aware of – their additional foreign citizenship. The circumstances suggest that was not intended.

This change was made only two weeks before the final session of the Convention ended (2 March 1898, 1744-5; 4 March 1898, 1914). At that time Barton and the Convention were under great pressure to resolve several controversial issues and bring a decade’s hard work to a successful close. The change was one of 140 amendments made by the drafting committee that Barton was asked to run through. He declined, suggesting that members instead review a page-by-page comparison with the previous draft. Barton portrayed the amendments as “merely questions of drafting” that did not alter the sense, except where required by the Convention (4 March 1898, 1914-5).

Other parts of s 44 were debated at length (7 March 1898, 1931-42), but there was no explanation or discussion of the dual citizen ban at all. It may be that the drafting committee did not realise the redraft would ban dual citizen parliamentarians. No one was seeking a dual citizen ban. There was no precedent for such a ban, and some of the federating colonies’ Parliaments probably had dual citizen parliamentarians. The colonies’ electorates that needed to approve the draft constitution would have included dual citizen voters.

It was more than two years after the Convention finished before the Constitution was enacted by the Imperial Parliament. During that time, the draft received close scrutiny by its opponents, the Parliaments and electorates of each federating colony, leading Australian and British lawyers, the British Colonial Office and the British government. Other changes were made as a result of that scrutiny, to secure support in NSW and at the insistence of the British government. However, no concern was raised about the late introduction of an unprecedented dual citizen ban that was contrary to the approach Britain itself had taken 30 years earlier and was about to reaffirm (1901 Report, [20]-[21]).

If the redraft was understood to create a dual citizen ban, the lack of discussion and opposition was surprising. It made perfect sense, however, if the 1900 approach applied.

As mentioned already, the 1900 approach meant the second limb of s 44(i) had limited effect, although it would still have limited how far Parliament could go if it removed the s 34(ii) requirement for parliamentarians to be subjects of the Queen. What the second limb would have prevented, without a referendum, was foreign citizen parliamentarians who owed no allegiance to the Queen. That was precisely what Barton had argued for at the Sydney Convention session (21 September 1897, 1013):

Persons who have taken the oath of allegiance to a foreign power are not to be classed in the same category as citizens of the country for the purpose of joining in legislation.

Barton’s statement accurately summarises the effect of the 1900 approach to s 44(i). Dual citizens, like any other citizens, would be disqualified under the first limb of s 44(i) if they swore allegiance to a foreign power. Dual citizens by descent and naturalised British subjects, who had not done that, would be treated the same as all other “citizens of the country”.  I suggest that Barton thought the redraft of s 44(i) was consistent with this.

I doubt that Barton was the only one who thought dual citizen British subjects were fit to be “joining in legislation”.  British subjecthood was widely considered a valuable privilege, giving those of foreign ancestry much-improved status.  They would not have been immune from prejudice (Chesterman 2005), but they were still British.

What would it mean for s 44(i) if the arguments above are correct?

At federation, the result of applying the 1900 approach to s 44(i) would have been:

  • Disqualification of parliamentarians did not turn on the vagaries of foreign laws and the procedures of foreign governments. Rather, it depended primarily on the first limb of s 44(i). The second limb had limited effect apart from preventing Parliament from permitting, without a referendum, the election of parliamentarians who were only foreign citizens and owed no allegiance to the Queen.
  • The test for disqualification was clear, based on matters likely to be within the knowledge of the person concerned, and did not distinguish between dual citizens by descent and British naturalisation. Application of the test to British subjects did not require foreign law advice, genealogical research or continuous monitoring of all foreign nationality laws.

That may help explain why s 44(i) did not appear to be problematic for over 90 years, including during two world wars when concern regarding allegiance was understandably extreme.

But what would this mean for s 44(i) today? Sykes v Cleary’s approach to the recognition of foreign law in determining dual nationality was said to be required by the common law and informed by principles of international law (at CLR 105-107, 110-112, 127-128, 131-132, 135, 139). If, as suggested above, the common law and international law on this were materially different in 1900, three things appear to follow:

  1. The 90’s approach resulted from a change in the common law, implying that the 1900 approach was not “entrenched” – since otherwise adoption of the 90’s approach would itself have involved altering the Constitution. Accordingly it may be possible for Parliament to further amend the common law on recognition of foreign law in determining dual nationality or restore it to the 1900 approach.
  2. It is possible that the original understanding of s 44(i) achieved its purpose so well, and for so long, that the assumptions underlying it were forgotten well before Sykes v Cleary. The High Court would be understandably reluctant to reconsider that decision. However, if the original reading of s 44(i) was, and is still, more certain and more consistent with its purpose – and if it has never been squarely put to the High Court – there is surely a strong case for that to occur. The historical context in 1900 was certainly raised in Re Canavan. See for example the submissions of the Attorney-General, the Amici Curiae, Senator Xenophon and the oral argument. However, those submissions did not ask the High Court to overrule Sykes v Cleary, and did not put the case that:
    • foreign nationality of British subjects was disregarded in the absence of clear contrary intent and
    • the second limb would not have been understood to apply, at all, to British dual citizens if the common law adopted its then usual approach of treating them only as British subjects and ignoring competing foreign claims to their allegiance.
  1. The 90’s approach created a problematic need for qualifications to the second limb of s 44(i) to give effect to the “constitutional imperative”. The 1900 approach had no need of that. It achieved its purpose, appropriately, on the basis of the plain meaning of the words as then understood. A British subject was British, not “foreign”, regardless of what foreign powers might claim.

A number of independent inquiries have considered s 44(i) and recommended removal or reform of the dual citizen ban. It may not be among the most important reforms to put to a referendum. Nevertheless, the ban is unfortunate and does harm. Our dual citizens should not be made to feel like second-class citizens. Doubting loyalty without reason undermines it. And this is not a good time to reduce unnecessarily the pool or diversity of eligible parliamentarians.

Bruce Dyer is a former Professor of Practice (Law) at Monash University. He is now principal of Conisante Consulting. A longer referenced version of this post and referenced legal arguments are available at

Suggested Citation: Bruce Dyer, ‘The dual citizen ban – what was Barton thinking?’ on AUSPUBLAW (27 September 2019) <—what-was-barton-thinking?/(opens in a new tab)>