Electoral reform and voters’ behaviour in Australia and New Zealand: editors’ introduction to the special issue

View correction statement:

Correction

As democratic political systems around the world face multiple challenges, from political polarisation to voter disengagement to loss of belief in the value of democracy itself, the topic of reforming democracy’s defining institution, elections, is of considerable importance. This special issue of Political Science presents five articles focused on electoral reforms in Australia and New Zealand. These are two of the world’s oldest and most stable democracies. Nevertheless, they are far from immune to contemporary global challenges to democracy. As parliamentary democracies sharing many similarities, they provide useful bases for comparison.

The articles address issues of early voting, majoritarian and proportional representation electoral systems, and citizens’ perceptions of effective representation. Three are selected from papers presented at the sixth annual conference of the Australian Society for Quantitative Political Science (ASQPS) held (fittingly for this special issue) in Wellington, New Zealand, in December 2017. This conference is now an established feature of political science in the Antipodes and the wider region, providing a forum for scholars using rigorous analysis of quantitative data to exchange ideas, approaches and findings. We hope this special issue will represent one further contribution stemming from these yearly gatherings.

The articles in this volume all focus on citizens’ behaviour in the context of particular electoral institutions, and in most cases, in the context of changes in those institutions. This kind of cross-temporal examination of important cases is one valuable way in which electoral reform can be studied, and contrasts with much of the existing literature on electoral institutions that takes a cross-sectional, cross-case approach. Specifically, two papers focus on before and after assessment of early voting (Garnett, Sheppard and Beauregard), one focuses on the change from majoritarian to proportional representation (Liu), another focuses on changes in a preferential voting system meant to limit the number of fringe parties gaining seats (McAllister and Muller) and one focuses on perceived effectiveness of representation in the context of lax campaign finance rules (Cameron and Wynter). All but the latter are able to take advantage of observed voter behaviour in the wake of institutional change, while Cameron and Wynter’s article provides a baseline against which behavioural change can be observed if and when Australia experiences significant campaign finance reform.

The papers draw on a broad array of data and approaches, analysing individual survey responses and electoral records. Several underscore the importance of repeat national election studies. Both the Australian (AES) and New Zealand (NZES) Election Studies provide crucial data over time that allow for temporal analysis of opinions concerning the perceived distribution of power in Australia (Cameron and Wynter) and voter decision-making in response to electoral reform in New Zealand (Lui). National election studies also typically contain rich content that allows for a close examination of the determinants of voter behaviour within a single election (Sheppard and Beauregard).

In this brief introductory essay, we attempt to synthesise some of the articles’ findings, provide an overview of the context of electoral reforms and briefly discuss the specific contributions of each article.

Considering the articles as a group, two aspects of their findings may be of particular interest: unintended and intended consequences of reforms,¹ and broader theoretical implications for political behaviour. Reforms that seem to have ‘worked’ include the introduction of early (‘advance’) voting in New Zealand in 2017 and the 2016 reforms to the way the Australian Senate is elected. Both changes were subject to intense public discussion and early voting is now an essential consideration in elections in Australia and New Zealand. Garnett provides evidence that increased turnout at the 2017 New Zealand election was linked to advance voting, and McAllister and Muller connect the reduced success of fringe parties in the Australian Senate to electoral reforms aimed at achieving just that. Sheppard and Beauregard also show that early voting in Australia appears to have reduced the costs of voting for some citizens, by reducing inconvenience. These are all important phenomena to understand as electoral commissions around the world think of how to better engage voters.

However, it is worth considering Sheppard and Beauregard’s findings in more detail. These show that early voters in Australia tend to be less engaged with politics, based on their scores on political knowledge questions in the Australian Election Study. They also point out that these early voters are not able to consider the full range of campaign statements and activity by the parties contending for their votes. This suggests their choices may not be as informed as those of voters who were exposed to the full campaign period up to election day, reducing the political engagement and knowledge of already disengaged voters.

One of the major differences between the Australian and New Zealand electoral institutions is Australia’s compulsory voting system. A great advantage of compulsory voting is that each citizen’s preference is counted (Fowler 2013). However, Sheppard and Beauregard’s findings suggest that early voting in this context may further exacerbate the problem of highly uninformed voters in such systems being easily influenced or making ‘poor’ choices based on irrelevant information (such as candidates’ smiles, see Horiuchi, Komatsu, and Nakaya 2012). It is very likely that political parties will start timing their messages to capture these early voters, but for now these electoral changes have created unintended consequences.

New Zealand provides more encouraging evidence on the democratic virtues of early voting. As Garnett demonstrates, voter turnout increased substantially in New Zealand following the introduction of no-excuse early voting in 2008. But she also uncovers other potential problems associated with early voting that differ from those revealed by Sheppard and Beauregard in Australia. In New Zealand, early voting is more prevalent in better-educated and urban electorates, raising concerns that it may skew the electorate further towards the affluent and well-informed. This is consistent with concerns raised by Berinsky (2006) that early voting may exacerbate existing socio-economic differences between voters and non-voters in the United States, leading to further inequities in electoral participation and political power.

The issue of the often-hypothesised connection between majoritarian electoral systems and the importance of leaders’ personalities in voter choice is taken up by Liu. She uses the 1996 New Zealand electoral reforms, moving from a majoritarian (first-past-the-post) to a mixed member proportional (MMP) system, to examine change over time in the importance of party-leaders’ images to voter choice. Surprisingly from the perspective of existing theory and literature, she does not find the expected reduction in the salience of party-leader image. Rather, it appears to increase after the reforms. This is a puzzling and intriguing finding. Can we find a causal connection? Liu cautiously concludes that New Zealand is simply not immune to the global trend towards personalisation of politics, even under proportional representation. She further suggests that this is to some extent independent of the electoral system. These are appropriately careful inferences made from the available data. But of course, Liu’s findings have wider implications and should invite further testing of the supposed causal link between electoral rules and the importance of parties and their platforms.

One area not explored very much in the articles is the motivations and processes giving rise to electoral reform. Fundamentally, reforms can be undertaken in order to strengthen democracy, or they can be undertaken to benefit one or several political parties or actors. Of course, in most cases, there will probably be some mixture of these motivations.² The case of the Australian senate reforms is perhaps most conspicuous in this light. They clearly benefitted the established parties, with their stated goal to reduce the chances of smaller parties entering the Senate.³ Conversely, the justification given by those supporting the reforms is also reasonable. Electoral systems that allow truly fringe parties to gain outsized influence distort democratic representation. Consistent with this reasoning, one of the major rationales for changing New Zealand’s electoral system was the disproportionality of election outcomes under the former system.

More generally, this distinction points to difficult but important issues that should be addressed to better understand the processes and outcomes of reforming electoral systems. It might be assumed that two (often related) necessary conditions need to be met to produce reforms. There must be a plausible argument that the reforms are needed to improve the functioning of the democratic system, and sufficient political support must exist for the reforms to be enacted. In the absence of the former, the changes are likely to be anti-democratic, and so not really ‘reforms’. In the absence of the latter, if the changes are somehow imposed – for example, by a bureaucracy that does not have the political mandate to do so or by outsiders – the situation can be one of ‘forcing them to be free’, or the imposition of democracy. This is self-contradictory, and perhaps unlikely to see success. It does appear that both of these conditions were met (to varying degrees) in the Australian and New Zealand cases presented here.

Considering cases in which both conditions are met – that is, cases of genuine electoral reform – there appears to be a problem of potential endogeneity in empirical analysis of outcomes. Reforms undertaken by political actors at a particular time in order to achieve certain ends might achieve those ends due to the actors’ motivations, rather than because of the general characteristics of the changes made. The same reforms undertaken in the absence of similarly motivated actors to implement them might lead to very different outcomes. This inferential problem is potentially a challenge for any observational study of institutional change, but it should not be overstated. If we have independent theoretical reasons to expect certain outcomes, regardless of implementing actors’ motivations, and the outcomes are consistent with that theoretical logic, concerns over endogeneity are reduced. If similar outcomes are observed in comparable cases, such concerns are further reduced.

Another telling aspect of the analyses in this special issue are the unintended consequences of reforms. These are highly unlikely to be the result of the political motivations of those enacting the reforms. So, they can speak more directly to the likely causal effects of the institutional changes themselves. Such may be the case with Australia’s early voting system – it seems unlikely that the legislators who passed the reforms anticipated or intended that it would further disengage voters from them during election campaigns.

An area – unlike the electoral reforms discussed above – where there is much less consensus between elites and the public is campaign finance. Like electoral system reform, campaign finance is a topic that is relevant to all established democracies. It is also one of the few areas in which Australia is a significant laggard, with very high reporting threshholds (as of late 2018, a donation does not need to be declared unless it exceeds AU$13,800), and long time lags between receipt of contributions and disclosure (up to 18 months). Cameron and Wynter find evidence that campaign finance laws may influence public attitudes towards politics and government, with Australia’s lax standards for election financing having potential consequences for the future health of the broader political system.

This brings us back to our initial point. Democratic political systems around the world face serious challenges. Reforms of the political system (or their absence) may help reduce (or exacerbate) citizens’ cynicism about political institutions and actors. The motivations behind, popular support for and impact of electoral reforms are key factors in this equation. In the case of campaign finance reform, Australia’s political elite appear to have placed their own interests above the preferences of their voters, and this has helped drive antipathy towards politics.

The articles featured in this special issue shed light on two important cases often overlooked in the international literature. Australia and New Zealand, among the world's most enduring representative democracies, have been innovative and successful, but they are not free from challenges. The findings in this issue of Political Science, we believe, help advance understanding of the causes and consequences of electoral system reform and should be of interest to scholars focused on these topics in democracies around the world.

Notes

1. We thank an anonymous reviewer for suggesting this focus.

2. We do not exclude the possibility that electoral ‘reforms’ may be designed to weaken democracy and benefit those in power by allowing them to retain power through non-democratic means. However, we believe such changes are of a qualitatively different nature than those discussed in these cases.

3. We thank an anonymous reviewer for suggesting this point.