The High Court Strikes Down a Campaign Finance Law (Again): Unions NSW v New South Wales

By Professor Adrienne Stone

Unions NSW v New South Wales Case Page

In 1992, in Australian Capital Television v Commonwealth [1992] HCA 1, the very first case on the constitutional freedom of political communication, the High Court struck down a Commonwealth law prohibiting electronic advertising during election periods. That law had been enacted as a campaign finance reform measure aimed at reducing the reliance of political parties on their donors and thus the High Court’s first application of the freedom of political communication struck a blow to the cause of campaign finance reform in Australia.

In more than twenty years since, however, freedom of political communication cases have focused on other questions such as the protection of political process, the application of defamation law in political debate and the permissibility of insult laws. It was not until late last year, however, that the Court returned to consider the operation of the freedom of political communication to the regulation of electoral finance. In Unions NSW v New South Wales [2013] HCA 58, the Court heard a challenge brought by unions to two sections of the Election Funding, Expenditure and Disclosures Act 1981 (NSW).

What were the challenged laws?
The general scheme of this Act requires disclosure of political donations by political parties, members of parliament, candidates and ‘third party campaigners’ (other persons who incur more than $2000 in electoral expenditure annually). It also caps the amount that can be donated to these persons and the total amount of electoral communication expenditure for State election campaigns.

The challenged provisions were inserted by amendment in 2012. The first, s 96D, prevented persons other than New South Wales electors from donating to New South Wales political campaigns. The second, s 95G(6) served to aggregate the electoral expenditure incurred by a party with ‘affiliated organisations’ for the purposes of determining the application of the expenditure caps.

An ‘affiliated organisation’ of a party is defined as a body ‘authorised under the rules of that party to appoint delegates to the governing body of that party or to participate in pre-selection of candidates for that party (or both)’. Having regard to party structures in Australia, the practical effect of this section is to ensure that expenditure by trade unions is aggregated with expenditure by the Australian Labor Party. Reflecting that fact, the challenge was brought by unions, each of which was authorised to appoint delegates to the annual conference of the New South Wales branch of the Australian Labor Party and who intended to make donations to it.

The High Court unanimously upheld the challenge in a relatively short judgment consisting of the joint reasons of French CJ, Hayne, Crennan, Kiefel and Bell JJ and the separate reasons of Keane J. (As described here, Gageler J did not participate.)

Like all cases concerning the freedom of political communication, the reasoning proceeds in two stages: first the Court asks whether the challenged law places an ‘effective burden’ on freedom of political communication and, if so, if it is ‘reasonably and appropriate and adapted’ to ‘a legitimate end compatible with the constitutionally prescribed system of representative and responsible government’.

The case turned principally on an element of the second of these questions: the question of whether the two challenged provisions could be characterised as a serving a ‘legitimate end’. But there were some important findings with respect to both questions.

Does the constitutional freedom of political communication apply to state politics?
One particularly welcome clarification was a clear statement of the principle that the freedom of political communication applies to political communication about state government and politics. This treatment of this question had a checkered history (discussed here). It was initially held, by way of obiter dicta, that Commonwealth and State politics were sufficiently entangled that the protection of the freedom applied to both forms of political communication. It appeared that position was modified as a result of Lange v Australian Broadcasting Corporation [1997] HCA 25.  In that case, the High Court was emphatic that the freedom of political communication operated to protect only institutions of representative and responsible government identifiable in the text of the Constitution rather than a more general concept of ‘representative government’. That position (addressed here) seemed to entail that discussion of state political matters were not covered by the freedom unless a specific link could be drawn to the operation of representative government at the federal level for instance by showing that the subject matter of the communication might affect voting in an election contemplated by ss 7 or 24 of the Australian Constitution.

Such a link might readily be drawn in some cases but the drawing of such a link could be by no means assumed. Hence in Levy v Victoria [1997] HCA 31 two judges found that a Victorian law regulating protest against duck hunting was not within the scope of the freedom because the subject matter of the protest (the Victorian government’s protection of duck hunting) had no connection to representative government at the federal level.

That position appeared to weaken over time, or at least it became clear that understood correctly it posed few if any limits on the concept of ‘political communication’. In Roberts v Bass [2002] HCA 57, Gaudron, McHugh and Gummow JJ described ‘statements made by electors or candidates or those working for a candidate, during an election, to electors in a State electorate, concerning the record and suitability of a candidate for election to a State Parliament’ as matters ‘at the heart of the freedom of communication protected by the Constitution’. In Coleman v Power [2004] HCA 39, which concerned complaints about the conduct of a member of the state police, the applicability of the freedom to the communication in question was a matter of concession.

In this case, the High Court seems to have settled the question for the time being. The Court recognised the increasing integration of economic social and political matters in Australia pointing out that our constitutional arrangements that contemplate and facilitate such integration such as the provision for Commonwealth funding of the states and cooperative legislative and executive schemes. In this light, the joint reasons, at [25], conclude that:

The complex interrelationship between levels of government, issues common to State and federal government and the levels at which political parties operate necessitate that a wide view be taken of the operation of the freedom of political communication.

The position seems to be then that the discussion of state political matters is considered to be covered by the freedom of political communication as a matter of course and there is no need to draw a specific link to matters relevant to the operation of federal institutions of representative government.

Indeed the joint reasons cite a passage from Lange in which a unanimous High Court stated that ‘each member of the Australian community has an interest in disseminating and receiving information, opinions and arguments concerning government and political matters that affect the people of Australia’. The import of the passage seems to be that the coverage of the freedom of political communication is determined in this single conception of ‘the government and political matters that affect the people of Australia’.

In declaring that the freedom of political communication covered the discussion of state political matters, the High Court also had to address a submission that the application of the freedom to the states was curtailed by the Melbourne Corporation principle. That is, in order to respect the constitutional independence of the states as recognised by the Melbourne Corporation principle, it was argued that the freedom of political communication should be applied to the states with a greater ‘margin for choice’.

The Court rejected this swiftly. In the words of the joint judgment at [34]:

the allowance of what is sometimes called the grant to the legislature of a margin of appreciation has been accepted by a majority of this Court. The latter is a large question and has not been seriously debated since the decision in Lange. In any event, there is a more fundamental difficulty with the argument which draws upon the Melbourne Corporation principle in an endeavour to quarantine State legislation from the operation of the freedom. There is no constitutional principle which accepts that the States can legislate to affect the Commonwealth Constitution, including its implications.

‘Is money speech?’
Having concluded that the freedom of political communication applied to communication about State political matters and that its application was unaltered by the Melbourne Corporation principle, the Court concluded that the law imposed an ‘effective burden’ on political communication.

In doing so, however, all members of the Court were sceptical about the relevance of plaintiff’s submission that ‘the making of a political donation is a form of political communication’. That submission show the influence of a tenet in US campaign finance law that ‘money is speech’, which has meant that campaign expenditures and campaign contributions have been regarded as ‘speech’ for purposes of the protection of the First Amendment.

But in Australia, as the Court reminds us, the source of the freedom of political communication lies in the underlying system of representative government. For Australian constitutional purposes, the question is not whether donation is political communication (the Australian equivalent of the question ‘is money speech?’) but whether the law burdens that system of government through its burden on communication.

Therefore to satisfy this element of the analysis, it is enough to show that the challenged laws ‘effect a restriction upon the funds available to political parties and candidates to meet the costs of political communication by restricting the source of those funds’. Both laws were found to impose this kind of a burden on political communication. In particular it was found that the burden imposed was not vitiated by the provision of public funding for election campaigns because it was apparent that parties would want to spend more than the amount provided by public funding and therefore parties would need to find a way to ‘fund the gap’ between public funding and the full costs of their campaigns.

Did the NSW laws have a legitimate end?
The Court then considered the second question: whether the impugned laws were ‘reasonably appropriate and adapted to a legitimate end’ … ‘compatible with the constitutionally prescribed system of representative and responsible government’.

At the core of the High Court’s reasoning was a finding that neither of the two challenged provisions could be characterised as serving a ‘legitimate end’. This finding was significant because freedom of political communication cases have rather more typically depended on the other element of the test: whether the challenged law was a ‘reasonably appropriate and adapted’ measure.

In relation to s 96D it seems that the Court was simply not convinced that there was any good reason for the enactment of the law. The Court accepted that the purpose of the overall scheme of the act was to prevent corruption or undue influence on the political process or the appearance of corruption or undue influence. The problem with s 96D was said to be that the Court could see no way that the exclusion of non-electors from the making of political donations advanced those purposes. In relation to corporations (who are included within the scope of s 96D), the joint reasons accept that there may well be reasons to have special concern about the role of corporations in financing campaigns but s 96D is not so limited. Moreover many of those excluded by s 96D ‘may have a legitimate interest in political matters’.

This conclusion — that the activities of non-electors might be protected by the freedom of political communication — is consistent with its underlying structure. As Keane J reminds us in his separate concurrence, the rationale for the freedom of political communication lies in its support of representative government. Its purpose is not to ensure that any particular persons or class of persons are able to engage in political communication but rather to ensure that all relevant communication is protected. Communication on matters relevant to the operation of representative government may well come from non-electors.

The problem with s 96D lay in a combination of over inclusiveness and under inclusiveness. The joint reasons acknowledge that political donations from some non-electors (such as, perhaps, corporations) could be limited in some circumstances but was not convinced that there was any reason to apply the prohibitions to such a large group of persons. In addition, the reasons contemplate that political donations might be prohibited altogether. It did not accept however that the category of non-electors could as a general matter be singled out in this way.

Section 95G(6) was found to be invalid for similar reasons. The joint reasons determined that the purpose of s 95G(6) was ‘to reduce the amount which a political party affiliated with industrial organisations may incur by way of electoral communication expenditure and likewise to limit the amount which may be spent by an affiliated industrial organisation’. The judges then concluded that the targeting of industrial organisations could not be connected to the ‘wider anti-corruption purposes’ of the Act.

The problematic aspect of the provision, then, was that it applied in effect only to a certain category of affiliated organisations — industrial affiliates — rather than affiliates in general. This aspect of the reasoning is even clearer in the judgment of Keane J. As his Honour put it, at [167]:

The effect of this differential treatment is to distort the free flow of political communication … Political communication generated by electoral communication expenditure by organisations affiliated with a party is disfavoured relative to political communication by entities which, though actively supportive of, and indeed entirely ad idem with, a given party, are not affiliated with it. To discriminate between sources of political communication in this way … is to distort the flow of political communication.

The case reveals much about the High Court’s approach to the freedom of political communication and the concept of representative government more generally.

Methodologically, the attention to the question of ‘legitimate end’ rather than the other aspects of the test (such as the ‘reasonably appropriate and adapted criterion’) appears to be something of a trend. It was the approach of the High Court in Coleman (explained here) and Monis v The Queen [2013] HCA 4 (explained here).

As a result of this trend, much depends on the process of statutory interpretation. Indeed how the Court characterises the end to which a law is directed might be determinative of its validity. The High Court’s decision last year in Monis makes the point. The High Court split 3:3 on the validity of a law directed at ‘menacing, harassing or offensive’ use of the post and the stance of the judges correlated strongly with the way they characterised the law in question. Those judges who characterised the law as directed to the ‘integrity of the post’ found the law invalid for want of a legitimate end. On the other hand those judges who upheld the law found that the law was directed to the prevention of very significant emotional harm attends the intrusion of the very offensive material into the home or workplace.

The spectre of the US
At a more philosophical level, the case gives some indication of the values underlying the freedom of political communication and the implication from representative government. In the minds of many people the spectre of US campaign finance law loomed large over this decision. Perhaps alarmingly, there are some elements of the High Court’s position this case that echo First Amendment positions. In particular the emergence of a non-discrimination norm evident in the High Court’s treatment of s 95G(7) appears to resemble the prohibition on content-based discrimination, which is perhaps the central tenet of First Amendment law.

This claim should not be overstated. There are many ways in which the law of freedom of political communication under the Australian Constitution differs from First Amendment law. Notably in this case, it is clearly contemplated that it might be possible to ban donations by corporations (which is not permitted under the First Amendment). But, this pattern of overtly rejecting, or at least doubting, the relevance of First Amendment law yet nonetheless converging of some of its positions is a familiar theme in this area of law. It was apparent in the Court’s early decisions on defamation and freedom of political communication (see here), in its approach to the law of insult (see here) and offensive letter writing (see here).

The source of this problem lies in way in which First Amendment law is typically distinguished by the Court. Usually, those judges who seek to make the point rely upon two features of the freedom of political communication: its structure as a limitation on power rather than a personal right and its underlying ‘institutional’ rationale of protecting representative government rather than the individual’s right to engage in ‘speech’. The first of these features is shared by the First Amendment which is also a negative, vertical right and which therefore provides no positive rights of action and only has application as between private individual where ‘state action’ is apparent. The second of these features is also insufficient to distinguish the freedom of political communication from the First Amendment because this ‘institutional’ rationale does not preclude consideration of personal values like the protection of an individual’s engagement in public debate. On the contrary, as argued here, there are some ways of understanding the relationship between political communication and representative government that would require strong protection of the individual’s capacity to engage in political debate.

Rather than relying on structural features, the High Court needs a clearer and more substantive vision of political communication under the Australian Constitution. In doing so it will need to decide to what extent parliaments are entitled to seek to ‘level’ the political playing field by limiting the political communication of the wealthy in order to enhance the voices of others. Given the continuing interest law reform in this area (see here and here) the chance to do so may come quickly.

AGLC3 Citation: Adrienne Stone, ‘The High Court Strikes Down a Campaign Finance Law (Again): Unions NSW v New South Wales’ on Opinions on High (14 May 2014) <>.

Professor Adrienne Stone is Director of the Centre for Comparative Constitutional Studies and Professor of Law at the Melbourne Law School.

3 thoughts on “The High Court Strikes Down a Campaign Finance Law (Again): Unions NSW v New South Wales

  1. Hi Adrienne, On the (topical) question of whether a narrower ban on unions/corporations would pass muster, I agree that the HCA’s express reasons leave that possibility open. But the problem is that corporations/unions are, in practice, the main example of non-electors who would want to make political donations. So, if the HCA thought that it would be OK to bar them from making donations, shouldn’t its ruling on s96D have been that s96D had a legitimate end (barring corps/unions from making donations) but that it wasn’t appropriate & adapted to that end?

  2. Jeremy, is the right way to look at it? The category of non-electors include persons with permanent residence and even non-residents. You are right that trade unions and corporations are particularly likely to (1) have money to donate and (2) have interests that make them want to donate but that might provide a reason for regulation rather than against. That is, the regulation is justified as it is aimed at those who are most likely to act in ways that distort or corrupt the system.

    As for the reasoning on s 96D, I think that there is little difference between

    (1) an argument that s 96 D is not adequately related to the legitimate end of addressing corruption and therefore reveals no legitimate end; and
    (2) an argument that s 96 D is not appropriately adapted to the legitimate end of addressing corruption (because over-broad).

    The High Court took the first path but it is substantially the same as the second I think.

    What do you think?

  3. As a matter of substance, I agree with you completely on both points. Anne Twomey at the G&T Constitutional Law Conference pointed out that one of the practical problems of s96D was that it made it difficult to seek street donations – e.g. Koala Bears with buckets – because it’s impossible to know who is on the roll. It’s a complete mystery to me why the NSW government didn’t introduce a narrower ban. Perhaps they somehow thought a broader ban would be more defensible, as the criteria would be electing, rather than ‘power’? Maybe the Greens were uncomfortable with being seen to target unions?

    But the HCA’s reasons is surprisingly silent on the problems on non-corporate non-electors and quite pointed in (for the first time) relying on legitimate end, rather than RA&A. (Anne Twomey suggested that this was a deliberate signal to Australian governments that they need to try harder.) So, reading the tea leaves, that suggests poor odds for a valid ban on corporates/unions (although it’s clear that there are good odds for a valid narrower ban on certain sorts of corporations, etc.)

    If a narrower ban was attempted (and, presumably, litigated), it would be an interesting instance of a ‘dialogue’ of sorts between the legislature and the HCA, although perhaps not a particularly transparent one.

Comments are closed.