Compulsory voting in Australia: Turnout with and without it

Compulsory voting in Australia: Turnout with and without it Jonathon Louth, University of Adelaide Lisa Hill, University of Adelaide ABSTRACT In this paper we address claims made by those supporting the abolition of compulsory voting about the relationship between turnout levels and compulsory voting. Via a critique of the methodology used, we query estimations of the effectiveness of compulsory voting laws and dispute common assertions about how high Australian turnout would be under a voluntary system. We then show that projected comparisons with places like Malta, New Zealand and The Netherlands are questionable. We also challenge other projections, that are based on data that has been insufficiently disaggregated. We conclude that when compulsory voting is properly administered in a congenial setting (such as Australia), it is the best means for guaranteeing high and socio-demographically equal rates of voting participation. Without it, Australian democracy would be experiencing the same citizenship crises currently being experienced in most other industrialised, voluntary voting settings. Jonathon Louth <[email protected]> is an Associate Lecturer and PhD candidate in the School of History and Politics at the University of Adelaide. His research and teaching interests are Australian electoral politics and international relations theory. He has a forthcoming book chapter ‘Coping with asymmetry: The complexity of the post-September 11 international system’, in The United States and the World beyond September 11 eds F. Patrikeeff and G. De Cure (forthcoming 2006). From 2003–2005 he was a research associate for The Compulsory Voting Project, and he recently won an APSA postgraduate travel award for a paper he submitted to this year’s annual conference. Lisa Hill <[email protected]> is a Senior ARC Fellow in the School of History and Politics at the University of Adelaide. Her interests are in political theory, intellectual history and issues in electoral law. She has published work on compulsory voting in Political Studies, Federal Law Review, Australian Journal of Political Science and Journal of Theoretical Politics (forthcoming). Lisa is currently working on a book about Adam Smith’s social thought and has just published The Passionate Society: The Social, Political and Moral Science of Adam Ferguson, Dordrecht: Springer/Kluwer, 2005. ISSN 1832-1526 Australian Review of Public Affairs Volume 6, Number 1: November 2005, 25–37 © 2005 School of Economics and Political Science, The University of Sydney 26 AUSTRALIAN REVIEW OF PUBLIC AFFAIRS Critics of Australia’s compulsory voting1 laws usually complain that such laws are an unjustifiable imposition on personal autonomy. In support of such claims, they often underplay the effect that compulsory voting has on turnout levels, their suggestion being that the cost to personal freedom is not justified by the gains to democracy. In this paper we address some of these claims and explore how Australia’s turnout levels might look under a voluntary system. By problematising the methodology used by compulsory voting critics and drawing attention to a number of constraining variables, we suggest that any optimism about likely turnout levels under a voluntary regime is misplaced. Background Australia has, it can be argued, the most efficient, effective and equitable compulsory voting system in the world and it enjoys turnout levels that are the envy of the industrialised voluntary voting world. It would be hard to find another system with such high levels of voter turnout coupled with such low transaction and opportunity costs to voters. Compulsory voting was introduced at the federal level in 1924 2 to address the problem of low voter turnout. It proved an extremely effective and well-tolerated remedy. The 1903 federal election turnout had only been 46.9 per cent of registered voters (RV) (Australian Electoral Commission [AEC] 1999), and at the last federal election held before the introduction of compulsory voting (1922) the average 3 turnout of registered voters had been 58.7 per cent.4 But turnout at the first federal election after 1924 (in 1925) surged dramatically, to an average5 of 91.4 per cent (RV). Data taken from the nine elections preceding and the nine following the introduction of compulsory voting shows that average voter turnout increased by 30.4 per cent (RV) (Jackman 2001). Turnout rates among the voting age population in Australia have remained consistently high, against the trend of steadily declining voter participation in advanced democracies (Blais 2000; Gray & Caul 2000). In the postwar period, the average turnout rate has been around 83 per cent of voting age 1 ‘Compulsory voting’ may be a misnomer because, due to the secret ballot, it is only registration (except in South Australia), recorded attendance and the acceptance of a ballot paper that are compulsory. But strictly speaking, marking the ballot formally and depositing it is also compulsory. For a fuller discussion of this point and any ambiguity surrounding it, see Hill 2002b. 2 The first state to introduce compulsory voting was Queensland (1914). Victoria followed in 1926, New South Wales and Tasmania in 1928, Western Australia in 1936 and South Australia in 1942 (McAllister et al. 1997, p. 71). But for the sake of simplicity, and in order to make meaningful comparisons with other systems, we restrict our attention here to first order (national) elections. 3 Broken down by house: 58 per cent for the Senate and 59.4 per cent for the House of Representatives. 4 Moreover, it had never been higher than 78 per cent (Irwin 1974, p. 293). 5 Senate: 91.3 per cent; House of Representatives: 91.4 per cent. LOUTH AND HILL: COMPULSORY VOTING 27 population (VAP) and 94.5 per cent (RV).6 In fact compulsory voting has been so effective at maintaining high turnout here that it has ‘rendered the study of turnout in Australian elections virtually irrelevant’ (Jackman 1997, p. 5). Other well-administered systems in established democratic settings have enjoyed similar results. For the period 1946–2003, Belgium has had an exemplary turnout rate, of around 93 per cent (RV). Similarly, in The Netherlands, for the 53 years when compulsory voting was in force (1917–70), turnout was consistently above the 90 per cent (RV) mark. In 1970, the Dutch Parliament voted 91–15 to repeal the legal compulsion to vote (Irwin 1974, pp. 292–93). A drop in turnout followed immediately. Between 1946 and 1967, the average turnout had stood at 94.7 per cent (RV). The overall average in the post-compulsory voting era dropped to 81.4 per cent—and it continues to decline. The last four national elections (1994–2003) showed an average turnout of 77.8 per cent (RV).7 From this data we can see that there has been a 12–15 per cent reduction in turnout (using both RV and VAP data) and as much as 20 per cent for individual elections.8 Importantly, while compulsion was in force in The Netherlands, it kept turnout above the 90 per cent mark for all socio-demographic groups. With its removal, an immediate consequence was an increased variation between subgroups (Irwin 1974, p. 294). This points to the long-asserted social ‘levelling’ effect of compulsory voting whereby all, rather than just the privileged and well-established sectors of society, are enabled to have their voices heard at election time.9 Compulsory voting thus seems to be a more than satisfactory solution to the problem of low and socially unequal turnout, and yet it is sometimes suggested that it is an inappropriate and unnecessary means of dealing with the turnout problem. In Australia, this has bolstered calls for its abolition. A voluntary voting Australia? Opponents of compulsory voting often assert that without the legal incentive Australians would still vote in large numbers. Fears that Australia’s turnout under a 6 For a discussion of the appropriateness of using RV or VAP data, see Hill and Louth (2004), where it is argued that when ‘considering the effectiveness of compulsory voting in (usually Western) industrialised states, RV is more appropriate when the issue is not to what extent the population may be disenfranchised but rather how effectively it mobilises entitled voters’. And further, that ‘any suggestions that Australia’s electoral system fails to mobilise a sufficient proportion of the population must be tempered by the fact that many people are legally disenfranchised’, as is the case for many permanent residents. 7 Calculated using data from International IDEA, at http://www.idea.int/. 8 At provincial elections held soon after the repeal, turnout was only 68.1 per cent—an immediate drop of over 25 percentage points. 9 For a discussion of compulsory voting as a social ‘leveller’, see Lijphart 1997 and Hill 2001. 28 AUSTRALIAN REVIEW OF PUBLIC AFFAIRS voluntary regime might be comparable to that of the US (normally in the low 50 per cent VAP range, though recently, and probably temporarily, somewhat higher) are dismissed as exaggerated (Minchin 1995, for example). Instead, compulsory voting critics insist that more realistic comparisons could be drawn from either New Zealand (Smithies 2004), Malta (Farrow 1997–98, p. 41) or The Netherlands (Minchin 2004, p. 30). These are three voluntary voting regimes that maintain fairly high turnout rates. We do not agree that these are meaningful comparisons for potential voluntary voter turnout figures in Australia. Misleading comparisons The use of Malta is particularly misleading, because it is far from representative. The fact that it often achieves the highest voter turnout in the world10 is due to the coexistence of an unusually large number of features known to be conducive to high turnout: a small, urbanised and geographically concentrated population (Siaroff & Merer 2002, p. 917); unitary, concentrated government; high levels of partisanship; proportional representation (PR); ‘highly competitive elections resulting in one-party governments despite P.R.’; extremely intense election campaigns; and a polarised electorate of partisan, committed voters (Hirczy 1995, p. 255). The New Zealand comparison is made more often, but again, it is questionable for a number of reasons. First, and eccentrically, 11 although voting is voluntary in New Zealand, enrolment is effectively compulsory (Bean 1998, p. 59). Were Australia to abolish compulsory voting but retain compulsory enrolment, it is doubtful that this would be enough to secure comparable voting levels; between 1911 and 1924 Australia actually did have compulsory enrolment, but turnout hovered in the low 50 per cent range. Second, New Zealand elections are more ‘salient’12 to the electorate because of the unitary and unicameral structure of government. Australia, on the other hand, is a federal and bicameral system (Jackman 1997, p. 409).13 In addition, any optimism 10 Average turnout for the last 7 elections in Malta (1971–98) (using all data available) is 95.3 per cent (RV) and 91.2 per cent (VAP). The most recent election (2003) saw a turnout of 95.7 per cent (RV). 11 New Zealand is a rarity among industrialised nations in upholding this practice in isolation from the compulsion to vote (Rose 1997, pp. 45–46). 12 ‘Salience’ is known to stimulate turnout. The ‘salience’ of elections denotes how significant, consequential, competitive, visible and meaningful they are to the electorate. For example, when elections are closely fought between distinctive parties and where the ‘the winner is empowered to put campaign promises into effect’, turnout will be higher (Franklin & Marsh 2002, p. 28). 13 New Zealand’s political culture is more intimate and less fragmented than Australia’s, where political identities are split and political focus divided. Unicameralism enhances the salience of elections because voting choices seem more consequential to voters in such systems. Undivided government (which New Zealand also enjoys) has a similar effect (Franklin & Hirczy 1998). Moreover, unicameralism is thought to reduce the information costs of voting (Gordon & Segura 1997, p. 137), thus removing yet another common disincentive to voting turnout. LOUTH AND HILL: COMPULSORY VOTING 29 about New Zealand turnouts should be dampened by the fact that the New Zealand turnout rate is currently in decline. 14 Even the recently introduced mixed-member proportional (MMP) system has not halted the general slide in turnout.15 In the seven elections held between 1981 and 1999 there was a cumulative drop of 12.8 percentage points (VAP). At the 2002 national election, turnout had slumped to its lowest postWorld War II point of 77 per cent (RV).16 Finally, even in the period before the introduction of compulsory voting in Australia, New Zealand’s turnout rate consistently outflanked Australia’s—by, on average, 20 per cent (Bean 1986, p. 71, n. 71). Some have suggested that under a voluntary regime, Australian levels might approximate those of The Netherlands, a country that abandoned compulsory voting in 1970. Since that time, although turnout has been respectable by industrialised voluntary voting standards, consistent and high turnout rates have been difficult to achieve. Between 1946 and 1967, the average turnout stood at 94.7 per cent (RV). After the removal of compulsory voting, the next seven elections (1971–89) averaged 83.5 per cent (RV). However, this figure disguises a gradual downward trend in turnout. The last four national elections (1994–2003) had an average turnout of 77.8 per cent (RV), and the overall average in the post-compulsory voting era is 81.4 per cent. What can be safely concluded from this data is that there has been a 12–15 per cent reduction in turnout (using both RV and VAP data), and as much as 20 per cent for individual elections.17 When the full effect of abandoning compulsory voting in The Netherlands became clear, Arend Lijphart, who had originally underestimated its importance (Lijphart 1968, p. 167) noted: ‘It is unlikely … that the Dutch would have made this decision [to drop compulsory voting] had they foreseen the disastrous plunge in their voter turnouts’ (Lijphart 1996). To describe such a plunge as ‘disastrous’ might be an overstatement. Many Australian compulsory voting abolitionists apparently regard turnout percentages in the high 70s as acceptable, if not good, but in the world of electoral studies, high 70s is a far cry from mid 90s. Further, we need to be careful about assuming that Australians would behave in the same manner as the Dutch, because some crucial elements of their political landscape are missing here. The Netherlands is geographically small, and has a more intimate political culture. It also has a unitary state and a full PR electoral system. These factors are generally thought to enhance electoral salience18 and therefore turnout.19 14 Average turnout for the 7 elections between 1981 and 1999 (using all data available) is 88.2 per cent (RV) and 82.2 per cent (VAP). 15 Though it did generate an immediate one-off spike in turnout. 16 VAP figures for this election were unavailable at the time of writing. 17 Calculated using data from International IDEA, at http://www.idea.int/. 18 See footnote 12. 19 For further discussion of the effects of PR on salience and therefore turnout (see Franklin 1999, p. 211; Hirzcy 1994, p. 65; Blais & Carty 1990). 30 AUSTRALIAN REVIEW OF PUBLIC AFFAIRS Reliability of survey estimates A number of other considerations also bear mentioning. Those who campaign for the abolition of compulsory voting often cite survey results which report that around 88 per cent of eligible Australians have indicated that they would most likely continue to vote under a voluntary system.20 Such results may be unreliable (i.e. inflated) for a number of reasons. To begin with, using a voluntary voting system, the turnout for the widely publicised election of delegates to the 1998 Constitutional Convention was only 46.6 per cent nation-wide.21 Further, as Simon Jackman has suggested, ‘surveybased estimates significantly under-estimate the extent to which turnout would decline under a voluntary turnout regime’ because of distortions generated by the response bias and the measurement error.22 These sorts of distortions are well known: a wide body of cross-national data indicates that survey respondents tend to overreport their voting participation, for instance (Katosh & Traugott 1981; Swaddle & Heath 1989; Anderson & Granberg 1997; Bingham Powell 1986, p. 27).23 This means that Australian turnout is likely to be much lower than surveys have indicated. Indeed, Jackman suggests that it could dip as low as levels ‘recorded in places like Japan and the US’: that is, to the 50–60 per cent VAP range (Jackman 1997, p. 42). It is doubtful, then, that the voting habit would stay with us without the strong incentive of law. Without compulsory voting, Australian democracy would be likely to be experiencing the same crisis of citizenship currently being experienced in most 20 An AGB McNair poll conducted in 1996 found that 88 per cent ‘of respondents indicated that they would be likely or very likely to vote’. The 1996 Australian Election Study arrived at roughly the same figure (Jackman 1997, p. 7). 21 Turnout varied nation-wide, from a high of 52 per cent in Victoria to a low of 40.3 per cent in the Northern Territory (Orr 1998, p. 578). Apart from its voluntary nature, it should be noted that other factors may have affected turnout at this election. First, none of the major parties ran campaigns; second, the vote was postal only (Orr 1998); and finally, it would probably rank as a second order rather than first order election, which means that its salience for voters was lower than for parliamentary elections. 22 The ‘measurement error’ refers to distortions caused by ‘social desirability effects’, whereby respondents answer in such a way ‘as to conform to societal norms’ rather than in terms of their actual intentions or past behaviour. In the case of the survey in question, the response bias works in the following way: because responding to surveys is voluntary and voting is not, the sample who have responded are more predisposed to vote anyway. This may lead to an overestimation of how many people in the general population will be prepared to turn up on election day (Jackman 1997). 23 This dynamic can occur whether subjects are reporting on past behaviour or on intended behaviour, as is the case in the above surveys: ‘Studies in the United States, Britain and Sweden have found that, probably out of a sense of what represents socially desirable behaviour, the public usually over-reports or exaggerates whether they had voted, when survey responses of reported behaviour are validated against the electoral register.’ Such results may be out by as much as 20 percentage points (Norris 2002, p. 84). LOUTH AND HILL: COMPULSORY VOTING 31 other advanced democracies, including New Zealand. Low and steadily declining voting levels are now a nearly universal phenomenon in industrialised, voluntary voting democracies worldwide. In a voluntary voting Australia, this crisis would undoubtedly be exacerbated by the introduction of a socio-economic status (SES) voting gap from which Australia has hitherto been largely immune. The inevitable reduction in turnout would, almost certainly, be concentrated among the more socially and economically marginal members of the electorate.24 Misuse of aggregate data We are wary of inferences drawn from single case studies, but we are just as wary of the obverse error: namely, the use of data that has been insufficiently disaggregated. The use of such data can lead to unwarranted optimism about how high Australian turnout would be without compulsory voting. Specifically, treating all compulsory voting regimes as a homogeneous group can give rise to an inaccurate (overly modest, in this case) appreciation of the performance of individual regimes such as Australia. We draw attention to the following example. Relying on aggregated statistics from the Institute for Democracy and Electoral Assistance (IDEA), the federal parliament’s Joint Standing Committee on Electoral Matters (JSCEM) report on the conduct of the 2001 election noted only a ‘small’ improvement in VAP voting rates using the mandatory vote. The report, quoting from IDEA, noted that ‘[A] somewhat surprising result of this study is that the 24 nations which have some element of compulsion associated with voting have only a small lead in turnout over the 147 nations without any compulsory voting’ (JSCEM 2003, p. 5).25 According to IDEA calculations, the mean VAP turnout for the compulsory voting group of nations stood at almost 70 per cent, and the non-compulsory group was only 7 percentage points behind, at 63 per cent. In effect, the results are distorted because Australia and other well-established, systematically administered compulsory voting states are grouped with states that have only nominal elements of compulsory voting, such as a brief reference in the constitution to the requirement of citizens to vote (Venezuela, for example). Many of the states identified as compulsory voting regimes either fail to enforce or lack the economic or institutional means to properly support compulsory voting. Some of these are new and semi-democracies in a state of economic development, places where good turnout levels may be hard to achieve 24 As is the case in almost all other industrialised voluntary voting settings where turnout is concentrated among the more settled, well off and better-educated members of the polity (Lijphart 1997; Hill 2002). 25 Nevertheless, the report did concede that ‘[o]ne reason for this is that the turnout figures we use are based on the total voting age population, not just on the number of persons enrolled to vote— where the compulsory voting countries do have a marked advantage—so that the impact of compulsory voting may only be significant if registration rates are also high’. See also IDEA 2002. 32 AUSTRALIAN REVIEW OF PUBLIC AFFAIRS Table 1. Voter turnout in three strictly enforced compulsory voting regimesa Time series electionsb Australia (AV)c 1983–98 Belgium (PR)d 1978–99 Netherlands (PR) 1946–67 Totals RV VAP RV VAP RV VAP RV VAP 1 94.6% 81.2% 94.9% 87.8% 93.1% 85.5% 94.2% 84.8% 2 94.2% 84.2% 94.6% 94.3% 93.7% 85.1% 94.1% 87.9% 3 93.8% 84.1% 93.6% 86.3% 95.0% 86.9% 94.1% 85.8% 4 95.3% 82.1% 93.4% 86.5% 95.5% 88.1% 94.7% 85.6% 5 95.8% 83.4% 92.7% 85.1% 95.6% 88.8% 94.7% 85.8% 6 95.8% 82.5% 91.2% 83.2% 95.1% 88.0% 94.0% 84.7% 7 95.2% 81.8% 90.6% 83.2% 95.0% 92.1% 93.6% 85.7% Average 95.0% 82.8% 93.0% 86.6% 94.7% 87.8% 94.2% 85.7% a The three countries represent properly institutionalised and strictly enforced compulsory voting supported by well-administered automatic or compulsory registration methods. They are also wealthy states (ranked by GDP per capita) and are all of a significant population size. Just as small states such as Malta and the Seychelles should not be used as examples of the potential of voluntary voting, compulsory regimes such as Luxembourg, Liechtenstein and Cyprus (to a lesser degree) are not fully representative of the potential of compulsory voting laws. They have thus been excluded. Further, neither Uruguay nor Cyprus has enjoyed a long enough period of electoral stability to be included in a seven election time series. Finally, the integrity of Uruguay’s registration system is questionable, and that of Cyprus is voluntary. b Time series are for the last seven elections when both RV and VAP figures have been recorded. c AV: Alternative Vote, based on a preferential majority system. Like the United States, the United Kingdom and Canada, Australia’s electoral system employs single-member districts. But it uses PR for its upper houses. d Although decline seems evident in Belgium, the most recent election (2003) saw a RV turnout of 96.4%, representing the highest turnout in the post-World War II era. VAP statistics are not yet available for that election. Source: International IDEA, available at http://www.idea.int/. LOUTH AND HILL: COMPULSORY VOTING 33 Table 2. Voter turnout in three weak or no enforcement compulsory voting regimes Time series electionsa Mexico (MMP) 1982–2000 Ecuador (PR) 1984–98 Bolivia (MMP)b 1980–2002 Totals RV VAP RV VAP RV VAP RV VAP 1 72.6% 63.8% 71.2% 56.1% ~ ~ 71.9% 60.0% 2 51.8% 45.1% 74.0% 64.9% 74.3% 59.1% 66.7% 56.4% 3 49.4% 41.4% 77.7% 67.0% 82.0% 65.2% 69.7% 57.9% 4 61.1% 50.0% 69.4% 64.7% 73.7% 51.0% 68.1% 55.2% 5 77.7% 65.9% 65.5% 66.3% 72.2% 50.0% 71.8% 60.7% 6 57.7% 54.4% 67.9% 67.8% 71.4% 64.5% 65.7% 62.2% 7 57.3% 48.2% 47.3% 48.5% 72.1% ~ 58.9% 48.4% Average 61.1% 52.7% 67.6% 62.2% 74.3% 58.0% 67.6% 57.6% a. Time series are for the last seven elections when both RV and VAP figures have been recorded. b. Election data for Bolivia before 1980 has not been used due to disturbances to the electoral process before that date. Source: International IDEA, available at http://www.idea.int. at the best of times; others impose suffrage restrictions that depress turnout. 26 According to our own reckoning, no more than 14 regimes out of a potential 30 can properly be described as being in any way compulsory—it is in fact rare to see the practice used with reasonable levels of enforcement and institutional support.27 This list can be further reduced to 4 (or 5 with the inclusion of The Netherlands until 1970) by restricting it to developed states with a history of well-established democratic norms and systematic administration of compulsory voting.28 26 For example, in Egypt voting is compulsory only for men. Further, undischarged bankrupts are prohibited from voting. Turnout hovers in the 30 per cent and below range. In Greece and Brazil voting is not compulsory for people over 70. 27 We recognise the following countries: Argentina, Australia, Belgium, Brazil, Costa Rica, Cyprus, Fiji, Greece, Liechtenstein, Luxembourg, Nauru, Peru, Singapore and Uruguay. The list can be extended to 19 countries if we include settings where the compulsion is not uniform—Austria (two Länder only) and Switzerland (one canton only)—or where the system was used for some time but later abandoned. Italy and Venezuela both dropped compulsory voting in 1993, and The Netherlands did so in 1970. 28 Systematic administration denotes systems where attendance is actively and universally encouraged and facilitated, avoidance is monitored and effective penalties are applied. For further discussion, see Hill 2004. These levels of systemisation are found are Australia, Belgium, Luxembourg, Cyprus and (before 1970) The Netherlands. 34 AUSTRALIAN REVIEW OF PUBLIC AFFAIRS Table 1 illustrates that when three established democracies with compulsory voting (and with good institutional support and enforcement) are grouped together, the average VAP turnout is almost 86 per cent, 23 percentage points higher than IDEA’s voluntary voting figure of 63 per cent and 16 points higher that its compulsory voting figure. The picture becomes even more clear when three developing compulsory voting regimes with little or no enforcement are grouped together (Table 2). Here the VAP turnout rate is just under 58 per cent—significantly less than the voluntary voting average—despite the fact that all of them use some variation of PR. Similarly, in semi-democratic compulsory voting regimes such as that of Singapore, levels of voting are actually lower than they are under comparable voluntary voting regimes. This highlights the pitfalls of using aggregate data that indiscriminately bundles together compulsory voting states that have vastly different levels of institutional support and democratic development. Concluding remarks Not everyone worries about low turnout. Some American voting libertarians have even suggested that low turnout is a positive, indicating satisfaction with government and politics.29 But compulsory voting abolitionists in Australia obviously disagree— their continued insistence that levels would remain relatively high and constant here under a voluntary regime implies that they perceive high turnout as something to be desired. Yet their optimism should be tempered by a number of conclusions that can be drawn from the above discussion. Inferences drawn from single or eccentric cases are likely to be suspect. So is any optimism fed by data that has been insufficiently disaggregated. It is indisputable that when compulsory voting is administered as well as it is in Australia, it is the most efficient and effective means for raising and maintaining high turnout. In fact, as Arend Lijphart has noted, compulsory voting is the only institutional mechanism that can achieve turnout rates of 90 per cent and above on its own (Lijphart 2001, p. 74). Although the effectiveness of compulsory voting is subject to variation (Hirczy 1994, p. 64), its efficacy should not be doubted (Franklin 1999, p. 206; Hirczy 1994, p. 64; Gordon & Segura 1997, p. 132). In an era when most industrialised democracies are battling to find solutions to the problem of declining voter participation, Australia continues to maintain high and steady voting levels. Without compulsory voting, Australian democracy would look very different. Turnout would be considerably lower and voting participation would be far less socially representative than it is at present. 29 Voter apathy has been described as a ‘political virtue’ that ‘may … have a beneficial effect on the tone of political life’ by providing an ‘effective counter-force to the fanatics who constitute the real danger to liberal democracy’ (Jones 1954, pp. 36–37). Russell Hardin has suggested that at the very least, low turnout is ‘evidence that government has not engendered grievous distrust and opposition’ (Hardin 1998, p. 24). LOUTH AND HILL: COMPULSORY VOTING 35 REFERENCES Andersson, H. E. & Granberg, D. 1997, ‘On the validity and reliability of self-reported vote: Validity without reliability’, Quality and Quantity, vol. 31, no. 2, pp. 127–40. Australian Electoral Commission (AEC) 1999, ‘The franchise “Right to vote” timetable’, media release, 16 September. Bean, C. 1986, ‘Electoral law, electoral behaviour and electoral outcomes: Australia and New Zealand compared’, Journal of Commonwealth and Comparative Politics, vol. 24, no. 1, pp. 57–73. Bingham Powell, G. 1986, ‘American voter turnout in comparative perspective’, American Political Science Review, vol. 80, no. 1, pp. 17–43. Blais A. 2000, To Vote or Not to Vote? The Merits and Limits of Rational Choice Theory, University of Pittsburgh Press, Pittsburgh. Blais, A. & Carty, R. K. 1990, ‘Does proportional representation foster turnout?’, European Journal of Political Research, vol. 18, pp. 167–81. Farrow, I. 1997, ‘Compulsory voting-the Australian anachronism’, Policy, Summer, pp. 41–43. Franklin, M. N. 1999, ‘Electoral engineering and cross-national turnout differences: What role for compulsory voting?’, British Journal of Political Science, vol. 29, no. 1, pp. 205–24. Franklin, M. & Hirczy, W. 1998, ‘Separated powers, divided government, and turnout in U.S. presidential elections’, American Journal of Political Science, vol. 42, no. 1, pp. 316–26. Gordon, S. B. & Segura, G. M. 1997, ‘Cross-national variation in the political sophistication of individuals: Capabilities or choice?’, Journal of Politics, vol. 59, no. 1, pp. 126–47. Gray, M. & Caul, M. 2000, ‘Declining voter turnout in advanced industrial democracies, 1950 to 1997’, Comparative Political Studies, vol. 33, no. 9, pp. 1091–122. Hardin, R. 1998, ‘Trust in government’, in Trust and Governance, eds V. Braithewaite & M. Levi, Russell Sage Foundation, New York. Hill, L. 2001, ‘“A great leveller”: Compulsory voting’, in Elections Full, Free & Fair, ed. M. Sawer, Federation Press, Sydney, pp. 129–41. Hill, L. 2002a, ‘On the reasonableness of compelling Australians to vote’, Political Studies, vol. 50, no. 1, pp. 80–101. Hill, L. 2002b, ‘Compulsory voting: Residual problems and potential solutions’, Australian Journal of Political Science, vol. 37, no. 3, pp. 437–55. Hill, L. 2004, ‘Compulsory voting in Australia: A basis for a “best practice” regime’, Federal Law Review, vol. 32, no. 3, pp. 479–97. Hill L. & Louth J. 2004, ‘Compulsory voting laws and turnout: Efficacy and appropriateness’, Proceedings of the Australasian Political Studies Association Conference, Adelaide, 29 September–1 October. Hirczy, W. 1994, ‘The impact of mandatory voting laws on turnout: A quasi-experimental approach’, Electoral Studies, vol. 13, no. 1, pp. 64–76. 36 AUSTRALIAN REVIEW OF PUBLIC AFFAIRS Hirczy, W. 1995, ‘Explaining near-universal turnout: The case of Malta’, European Journal of Political Research, vol. 27, no. 2, pp. 255–72. International Institute for Democracy and Electoral Assistance 1997, Voter Turnout from 1945 to 1997: A Global Report on Political Participation, International IDEA, Stockholm [Online] Available: http://www.idea.int/97df/presskit/turnout.html [1999, Jun 5]. International Institute for Democracy and Electoral Assistance, 1999, Voter Turnout from 1945 to Date: A Global Report on Political Participation, International IDEA, Stockholm [Online] Available: http://www.idea.int/Voter_turnout/parliamentary.html [2002, Apr 19]. International Institute for Democracy and Electoral Assistance, Voter Turnout Since 1945: A Global Report, International IDEA, Stockholm [Online], Available: http://www.idea.int/vt /survey/voter_turnout8.cfm [2004, May 5]. Irwin, G. 1974, ‘Compulsory voting legislation: Impact on voter turnout in The Netherlands’, Comparative Political Studies, vol. 7, no. 3, pp. 292–315. Jackman, R. W. 1987, ‘Political institutions and voter turnout in the industrial democracies’, American Political Science Review, vol. 81, no. 2, pp. 405–23. Jackman, S. 1997, ‘Non-compulsory voting in Australia?: What surveys can (and can’t) tell us’, Political Methodology Working Papers 1997 [Online], Available: http://wizard.ucr.edu/polmeth /working_papers97/working_papers97_alpha.html [Oct 2001]. Jackman, S. 2001, ‘Voting: Compulsory’, in International Encyclopedia of the Social & Behavioral Sciences, editors-in-chief N. J. Smelser & P. B. Baltes, Elsevier Science, Oxford. Joint Standing Committee on Electoral Matters 2003, The 2001 Federal Election: Report of the Inquiry into the Conduct of the 2001 Federal Election, and Matters Related Thereto, Parliament of the Commonwealth of Australia, Canberra. Jones, W. H. Morris 1954, ‘In defence of apathy: Some doubts on the duty to vote’, Political Studies, vol. 2, no. 1, pp. 25–37. Katosh, J. P & Traugott, M. W. 1981, ‘Consequences of validated and self-reported voting measures’, Public Opinion Quarterly, vol. 45, no. 4, pp. 519–35. Lijphart, A. 2001, ‘Compulsory voting is the best way to keep democracy strong’, in Points of View, eds R. E. DiClerico & A. S. Hammock, McGraw-Hill, New York. Lijphart, A. 1997, ‘Unequal participation: Democracy’s unresolved dilemma’, American Political Science Review, vol. 19, no. 1, pp. 1–14. Lijphart, A. 1996, ‘Compulsory voting is the best way to keep democracy strong’, The Chronicle of Higher Education, vol. 43, no. 8, pp. B3–4. Lijphart, A. 1968, The Politics of Accommodation: Pluralism and Democracy in The Netherlands, University of California Press, Berkeley and Los Angeles. McAllister, I, Mackerras, M. & Boldiston, C. B. 1997, Australian Political Facts, Macmillan, Melbourne. Minchin, N. 2004, Counterpoint, Radio National, 24 May [Online], Available: http://www.abc.net.au/rn/talks/counterpoint/stories/s1117921.htm [2004, Dec 5]. LOUTH AND HILL: COMPULSORY VOTING 37 Minchin, Nick 1995, ‘Rights, responsibilities and the electoral system’ [Online], Available: http://home.vicnet.net.au/~victorp/liberals/sa/minchin/html [1999, Sep 28]. Norris, P. 2002, Democratic Phoenix: Reinventing Political Activism, Cambridge University Press, Cambridge. Orr, G. 1998, ‘Tinkering with convention: Voluntary voting at Australia’s Constitutional Convention’, Electoral Studies, vol. 17, no. 4, pp. 575–81. Rose, R. 1997, ‘Evaluating election turnout’, in International Institute for Democracy and Electoral Assistance, 1997, Voter Turnout from 1945 to 1997: A Global Report on Political Participation, International IDEA, Stockholm [Online], Available: http://www.idea.int/97df /presskit/turnout.html [1999, Jun 5]. Siaroff, A. & Merer, J. 1990, ‘Parliamentary election turnout in Europe since 1990’, Political Studies, vol. 50, no. 5, pp. 916–27. Sigelman, L. 1982, ‘The non voting voter in voting research’, American Journal of Political Science, vol. 26, no. 1, pp. 47–56. Smithies, B. 2004, Counterpoint, Radio National, 24 May, [Online], Available: http://www.abc.net.au/rn/talks/counterpoint/stories/s1117921.htm [2004, May 28]. Swaddle, K. & Heath, A. 1989, ‘Official and reported turnout in the British general election of 1987’, British Journal of Political Science, vol. 19, no. 4, pp. 537–51.