“The type of issue that could be canvassed under Section 51 of the constitution — simply at the moment, in Clause 21, it just says ‘marriage’,” Mr Morrison said. “You could equally put in there opposite- and same-sex marriage and clarify very clearly what the meaning of the constitution is on this question, and to reflect [what] some would argue has been a societal change since the constitution was first written.”

Mr Morrison acknowledged the High Court had already ruled on it. “Justices of the High Court have already expressed opinions on this issue, that’s fine, but what I am saying is I would prefer the Australian people decide this: not me, not [High Court Chief Justice Robert French], but the Australian people.”

Federal Minister for Social Services Scott Morrison here refers to Cth v ACT [2013] HCA 55, where six members of the Court said that ‘When used in s 51(xxi)’ of the Constitution, the federal Parliament’s power to make laws about marriage, ‘”marriage” is a term which includes a marriage between persons of the same sex.’ Attorney-General George Brandis later relied on the same case to declare that ‘No constitutional referendum is necessary in this case.’

Given the High Court’s 2013 holding, what would be the legal effect of the referendum? There are two possibilities to consider. One possibility is that the referendum will succeed, writing the view of six High Court judges in 2013 permanently into the Constitution. While that won’t change the law, it will have the effect of barring a future High Court from disagreeing with that particular holding. Specifically, it would remove the power to decide from four future High Court judges, for instance stopping Gageler, Nettle and Gordon JJ (none of whom participated in the 2013 decision) from getting together with French CJ’s successor sometime after 2017 to rule that the federal parliament lacks power to enact a same-sex marriage law (effectively putting the political ball in the court of state or territory parliaments.) It may also make it less likely that a future High Court will overturn the central holding in Cth v ACT that territory (and, almost certainly, state) laws for same-sex marriage cannot operate unless the terms of the federal marriage law is changed, because such a ruling would arguably be inconsistent with the apparent intent behind the referendum of leaving that question exclusively in the hands of the federal parliament. The same would go for the already very unlikely possibility that a future High Court would hold that the federal parliament cannot constitutionally exclude same-sex marriages from its marriage law. For past examples of the Court examining the debates preceding a referendum in order to interpret a resulting constitutional provision, see here and here.

But what about the other possibility (one presumably hoped for by many of the referendum’s current proponents), that the referendum will fail (either by failing to attract a majority of Australian voters, or failing to attract a majority of voters in at least four states)? That would leave the Constitution unchanged, but could it affect a future High Court’s willingness to revisit its earlier rulings (e.g. on the basis that the referendum signals that the Australian people disagree with the 2013 ruling?) In a 1997 case on whether territory governments could acquire property without just terms, Gaudron J and Kirby J split on whether the fact that a majority of ACT residents voted against a 1988 referendum on this issue could be taken into account. In 2006, a majority of the High Court firmly rejected relying on failed referenda in the decision upholding the Howard government’s workplace relations law, in these terms:

There are insuperable difficulties in arguing from the failure of a proposal for constitutional amendment to any conclusion about the Constitution’s meaning. First, there is a problem of equivalence. The argument must assume that the proposal which was defeated was as confined as is the question that now falls for determination. If the proposal was wider than the immediate question for decision, it is not open to conclude that a majority of those to whom the proposal was put (whether they are described as “the people of Australia”, the “sovereign force” or, as in s 128, “the electors qualified to vote for the election of members of the House of Representatives”) reached any view about the ambit of the (unamended) constitutional power, or that they reached any view about that part of the proposal that appears to deal with the immediate issue…

Secondly, despite Harrison Moore’s optimistic view that the constitutional alteration mechanism provided by s 128 was a “less cumbrous” way for avoiding the obstacle of disagreement between the Houses of Parliament than the deadlock provisions of s 57 of the Constitution, few referendums have succeeded. It is altogether too simple to treat each of those rejections as the informed choice of electors between clearly identified constitutional alternatives. The truth of the matter is much more complex than that. For example, party politics is of no little consequence to the outcome of any referendum proposal. And much may turn upon the way in which the proposal is put and considered in the course of public debate about it. Yet it is suggested that failure of the referendum casts light on the meaning of the Constitution.

Finally, is the rejection of the proposal to be taken as confirming what is and always has been the meaning of the Constitution, or is it said that it works some change of meaning? If it is the former, what exactly is the use that is being made of the failed proposal? If it is the latter, how is that done?…

Constitutional construction is not so simple a process as the argument from failed referendums would have it. If, as is so often the case, a question about the meaning and operation of the Constitution is controversial, it is for this Court to determine the answer that is to be given… The constitutional text must be treated as the one instrument of federal government.

Assuming a future court agrees, this ruling implies that a failed referendum on same-sex marriage would have no legal effect at all on how that issue is eventually resolved. Of course, the referendum could well have a political effect, not only on politicians, but also, perhaps, on the willingness of Australia’s judges to issue holdings that differ from a clearly expressed public vote.