At the recent British election, Tony Blair won a third term in office. The turnout was some 61.28% of the electorate and, off 35.2% of the vote, Labour won 55% of the seats. Twenty-one percent of the total electorate delivered a historic win to New Labour. Some Australian commentators were appalled. A government elected with so little support couldn’t possibly be legitimate. Senator Andrew Bartlett, for example, wrote, ‘the Prime Minister’s ability to hold power with the support of just a fifth of the British adult population is the lowest figure since the Great Reform Act of 1832 … Britain [has] a 19th Century voting system.’[1] As almost every news story of the UK election told us, Britain doesn’t have compulsory voting. As if this were somehow unusual. In fact, Australia is the outlier. It is extremely unusual for an industrialised, advanced democracy to practice compulsory voting. Australia is just one of 24 countries that have the practice.[2] Even this number, however, is misleading. Of those countries, not all have a penalty for non-voting. Others have compulsory voting for some elections only, while Egypt has compulsory voting for men only.
Compulsory voting was first introduced in Queensland in the 1915 election. A conservative government, under the impression the ALP was using shop stewards to get out the vote, made voting compulsory—but lost the subsequent election anyway. The Commonwealth introduced compulsory enrolment in 1911, and compulsory voting in 1924.[3] There was little controversy at the time. Political parties benefited, as they still do, because they did not need to waste resources on a ‘get out the vote’ campaign.[4] Compulsory voting is a wealth transfer from those individuals who would not vote, and from the Australian Electoral Commission (AEC), which expends substantial sums of money tracking down individuals in order to enrol them to vote, to political parties.
The AEC list a series of arguments for, and against, compulsory voting at their website. In fairness to the AEC, they are simply charged with enforcing the electoral laws and do not make those laws themselves. The AEC lists six arguments in favour of compulsory voting. The first is that voting is a civic duty similar to paying tax, serving jury duty and the like. Compulsory voting is said to teach voters the benefits of political participation. Furthermore, compulsory voting requires the government to consider the total electorate, and ensures the election reflects the will of the people. Political parties can also concentrate of the important issues of the day, rather than simply ‘getting out the vote’. Compulsory voting ensures that the vote more accurately reflects the will of the electorate. Finally, the argument is that voters are not actually required to vote, but simply to arrive at the polling place and have their name struck off the voters’ roll.
In this article, we investigate the arguments in favour of compulsory voting. Specifically, we provide empirical analysis, as opposed to philosophical retort. We confront the ennobling ideals of compulsory voting with the harsh reality of judicial comment. We explore the notion that compulsory voting requires political parties to consider the entire electorate. Finally, we investigate the outcomes of compulsory voting with an estimated voluntary result. The findings are not pretty. The arguments in support of compulsory voting do not fare well. Those arguments that do fare well reflect poorly on the political process.
The duty to vote
Section 245 of the Commonwealth Electoral Act 1918 deals with compulsory voting. Indeed, that section of the Act is entitled ‘Compulsory Voting’. Subsection (1) of s245 states, ‘It shall be the duty of every elector to vote at each election.’ The remaining 17 subsections all deal with non-compliance. For example, subsection 9(d) indicates the fine for not voting is $20, while subsection 15 states the non-voter is fined $50, if they go to court. Non-voters may be excused the fine if they provide a ‘valid and sufficient reason for failing to vote’, or if they are dead, ineligible to vote, or were absent from Australia on polling day (subsection 4). The penalty for providing misleading, or false, information concerning the failure to vote is $50. According to the AEC, 52,796 non-voters paid the $20 fine after the 2004 election—an increase from 39,874 after the 2001 election.[5]
Despite the plain language of the Act, many Australians are under the impression that voting per se is not compulsory. Rather it is compulsory to simply turn up to the voting booth, and have your name struck off the register. Whenever compulsory voting is discussed in the media, the letters pages of the newspapers are filled with correspondents claiming voting is not compulsory, while attendance is compulsory.[6] Before discussing this notion in greater detail, consider the case of Krosch v Springell: ex parte Krosch [1974] QdR 107. Mr Springell arrived at a polling booth and handed the electoral officer a note saying he did not wish to vote for any candidate, as he found them all to be unworthy. He was prosecuted, and fined, for not voting despite the fact he had made the effort to turn up on Election Day. In its submission to the Joint Standing Committee on Electoral Matters (JSCEM) Inquiry into the 2004 election, the AEC makes the claim ‘Because voting is compulsory in Australia, turnout is regularly in the vicinity of 95%.’[7] Clearly, the AEC takes the view that compulsory voting leads to high turnout, not compulsory turnout leads to high voting.
Colin Hughes, the Australian Electoral Commissioner from 1984–89, has argued that voters were not compelled to vote for any candidate, they could always spoil their ballot.[8] Kim Beazley, leader of the opposition, has argued voters can simply put the ballot in their pocket and leave the polling booth (in apparent defiance of section 339(1) that explicitly defines this behaviour as an offence). While Chris Puplick AM, former Liberal senator, has written that voters ‘have an absolute right not to vote by placing a blank or spoiled ballot paper in the ballot box’.[9] Judges enforcing the law, however, appear to be unaware of this ‘absolute right’, especially with regard to blank votes. Chief Justice Barwick, for example, wrote that voters must actually mark the ballot paper and deposit that ballot into a ballot box. Justice Blackburn was of the opinion that casting an invalid vote was a violation of the Act. Both views are consistent with section 233 of the Electoral Act where the process of voting is described.
Voters, increasingly, do spoil their ballots (see Figure 1). This, of course, begs the question, ‘What does happen to individuals who vote informally?’ Nothing. Australia was the first country to employ the secret ballot (often referred to as the ‘Australian Ballot’). According the Australian Electoral Commission, it is not ‘an offence to vote informally in a federal election.’[10] After all, there is a secret ballot (section 233(1) of the Electoral Act); in principle the authorities should be unable to identify any voter from his or her actual ballot. For example, any ballot where the voter has written his or her name is invalid. Furthermore, many voters may ‘legitimately’ spoil their ballots through confusion and error. This view, however, is not entirely consistent with an AEC Research Report where the author writes, ‘While compulsory voting avoids a high degree of abstention, there is no guarantee that everyone will comply with the electoral laws and vote formally’ (emphasis added).[11] Similarly, the JSCEM report into the 2004 election states, ‘Because of the secrecy of the ballot, it is not possible to determine whether a person has filled out their ballot paper prior to placing it in the ballot box. It is therefore not possible to determine whether all electors have met their legislated duty to vote’ (emphasis added).[12]
This inconsistency may well explain the confusion.[13] Voters are legally required to vote, but there is no penalty for spoiling your ballot. Indeed, given that failing to vote because the voter dislikes all candidates equally, or cannot decide among them, is not a valid and sufficient for not voting, many voters may well have to spoil their ballots, or pay a fine. A reading of some of the court cases involving compulsory voting is instructive. The definitive case is Judd v McKeon (1926) 38 CLR 380 where the general principle of ‘valid and sufficient’ was determined to be a ‘personal physical inability to record a vote’. Of course, the court left open the notion that ‘valid and sufficient’ cause was a function of the circumstances, yet it is difficult to understand what these other circumstances could be. They are not, for example, an ideological objection to voting (Judd v McKeon), nor is a lack of preference for any candidate (Faderson v Bridger 1971 126 CLR 271), or even ignorance of the candidates (O’Brien v Warden 1981 37 ACTR 13). The latter cases are quite remarkable. Mr Faderson indicated he had no preference for any of the candidates, and to say he had would constitute a lie. High Court Chief Justice Barwick indicated that voters are not expected to express an opinion of what they want, but merely to indicate, from the choice available to them, what they must have. This neatly sidesteps the issue of voters being forced to lie. The Warden case is even more damning. Here the issue of voters lying is not sidestepped. Mr Warden arrived in the ACT just prior to an election, and was ignorant of the candidates and their policy platforms. Nonetheless, he was found not to have a ‘valid and sufficient’ reason for not voting. In the words of Chief Justice Blackburn of the ACT Supreme Court: ‘In my opinion the Act does not oblige the elector to make a true expression of his preference among the candidates. On one view he must make an expression of apparent preference; on another he need not express himself intelligibly or at all.’
Judges interpret the law as it stands. It is difficult, however, the reconcile the legal judgements reviewed here, with the more noble propositions that compulsory voting reflects the will of the people, or that it teaches the benefits of political participation. Voters are taught, ‘he need not express himself intelligently or at all.’
At present, voting technology is such that the spoilt ballot ‘anomaly’ survives. In the privacy of the voting booth, with pencil and paper any number of legitimate errors or confusions can give rise to a spoilt ballot. In future, however, with computerised voting technology, it will become impossible to cast a spoiled vote. Already commercial online registration forms do not allow the user to proceed until all boxes are appropriately ticked and information provided. Given the wording of the Electoral Act, the AEC are well within their rights to design electronic voting systems that require a valid ballot before the voter’s name is marked as having voted. Furthermore, electronic safeguards can be designed such that voting behaviour and voter identity cannot be matched. At that point the secret ballot will no longer protect those who purposefully spoil their ballots and so circumvent section 245 of the Electoral Act.
Considering the total electorate
Luke Andrews, Tim Fry and Keith Jakee investigate the economic properties of a compulsory voting regime relative to voluntary voting regime.[14] Their model makes some powerful predictions. It is easier to identity marginal electorates under a compulsory voting regime; therefore we can expect greater pork barrelling in those electorates and less in safer electorates relative to a voluntary voting regime. They test their hypothesis for Australia by investigating the Job Network program. Consistent with their hypothesis they find, every being equal, greater Job Network resources in marginal electorates. Similarly, Andrew Leigh found greater expenditure in the Roads to Recovery program in seats that swung to the government at the 2004 election.[15] These findings are inconsistent with the notion that governments must consider the entire electorate. Compulsory voting reduces the costs to political parties of getting out the vote and allows them to concentrate resources in marginal electorates.
How much lower would turnout be under voluntary voting?
We have data from the Australian Election Survey (AES) conducted after each federal election that asks whether or not the respondent would have voted if voting had not been compulsory.[16] On the surface, this is a guide to turnout under a voluntary regime. There is, however, an upward bias in the AES data. Simon Jackman has indicated that the type of person willing to participate in the AES is more likely to be interested in politics and consequently more likely to vote anyway.[17] Actual turnout under a voluntary voting system would be lower than the AES indicates. Jackman derives a weighting scheme to estimate unbiased vote shares for each political party. Using the 1996 House of Representatives election data, he shows that the Coalition would benefit from a change to voluntary voting. The ALP and minor parties, other than the Democrats, are worse off under a voluntary voting regime. The Democrats may or may not be worse off as the evidence for them is mixed. Jackman, however, finds that the impact of voluntary voting, while favouring the conservative side of politics, is not as great as is commonly believed.
We replicate the Jackman procedure for all elections since 1996 for both the House of Representatives and the Senate.[18] This is an important difference from previous research into compulsory voting. The general consensus is that compulsory voting benefits the ALP and minor parties,[19] but that research has only investigated voting in the House of Representatives, not the Senate. Unfortunately, the data required for these estimates do not exist prior to 1996. Consequently, we are unable to state whether our results favour the Coalition, or the government of the day. We also calculate two measures of our own. The first voting measure (Version 1) is the intersection of those individuals who believe voting should be compulsory, and who would definitely vote if voting were voluntary. Our second measure (Version 2) consists of those individuals who think voting should be compulsory, and who would either definitely or probably vote if voting were voluntary. In essence the view that voting be compulsory can be thought of as a measure of civic duty, indicating that such people are more likely to feel obliged to vote. Of course, there are many individuals who would vote even though they don’t believe voting should be compulsory.
We show the estimated turnout in figure two, comparing it to the actual turnout, which is very high, reflecting Australia’s compulsory voting regime. The Jackman score declines consistently over the last four elections. This is due largely to the construction of the measure itself – the AES response rate is declining over time.
All the adjusted turnout figures are far lower than the current actual, and the AES score. The adjusted turnout estimates range from 55% to 69% (the British turnout was 61%). This is unsurprising. What is surprising is the extent of the gap between the actual turnout and the estimated turn out. Arend Lijphart,[20] in his extensive survey of compulsory voting, indicated a gap of the order of 7–16% or nearly half that shown here. On the other hand, our Version 1 measure is very restrictive, and is likely to represent a lower bound of turnout figures.
Would voluntary voting have affected election results?
The estimated vote shares for political parties are shown in table one.[21] In each of the last four elections, the Liberal Party would have had a higher vote share in the House of Representatives under a voluntary voting regime. In two of the last four elections (1998 and 2004), the Jackman adjustment predicts the ALP would also have had a higher vote share. In addition, our two measures also predict a higher vote share for the ALP in two elections (1998 and 2001). Despite the higher vote shares for the ALP, it is unlikely they could have won three of the last four elections. Clearly, in 1996 the ALP lost government in a landslide, while in 2001 and 2004 the combined Coalition vote easily dominates the ALP vote, despite the increases. The ALP won the popular vote in the 1998 election, but not enough seats in the parliament. We cannot say with certainty that the Coalition would have won government in that year under a voluntary voting scheme. We are inclined to the view, however, that the Coalition would have won office under a voluntary voting regime. Looking at the actual vote share, we see the combined Coalition vote was less than the ALP vote. In each of the adjustments, however, the combined Coalition vote is greater than the ALP vote. Our conclusion, looking at the House of Representatives, is that the same government would be elected under either a compulsory voting regime or a voluntary voting regime.
Winning government and implementing policy is not quite the same thing. Legislation needs to pass the Senate. In each of the last four elections, the Coalition would have had a far higher vote share under a voluntary regime than it does under the current compulsory voting regime. Similarly, in each election, the ALP would have had a smaller vote share. The increase in Coalition vote share is not at the expense of the ALP—in some years the ALP decline is not statistically significant. Their vote share declines by a small amount. The biggest decline in vote share occurs in informal voting, One Nation and Independents. The decline in informal voting is quite remarkable. This result is consistent with a submission to JSCEM 2004 by Antony Green. He argued, ‘deliberately spoilt ballot papers make up only a minority of the informal vote.’ Most informal ballots, he argued, are due to incorrect or incomplete marking.[22] Our analysis indicates that those individuals who don’t understand how to engage with compulsory preferential voting would not vote under a voluntary system. In other words, an additional cost of compulsory voting regimes is to force individuals to the ballot box who are too embarrassed to ask for assistance, and who otherwise would not vote.
In each of the last four elections, the Democrats and the Greens would have had a higher vote share if there had been a voluntary voting regime. The Democrats are huge losers under the current compulsory regime. In 1998, they may have received as much as 15.63% of the Senate vote, as opposed to the actual 8.18% they did receive. The composition of the Senate would be different. There would have been fewer independents, and fewer ALP Senators. There would have been more Democrats, more Greens, and more Coalition Senators. After the 1996 election, the Coalition may well have had control over the Senate, however, based on the 1998 election results, they may well have lost that control. Overall, however, we cannot state unambiguously that the Coalition would have had absolute control of the Senate. We can say there would have been greater representation of left-leaning minor parties, and less representation for One Nation and independents.[23]
Conclusion
A remarkable component of Australia’s compulsory voting regime is the denial: supporters of compulsory voting deny its existence. Kim Beazley has said, ‘We don’t have compulsory voting in Australia. … The people who go out and put this claptrap argument up, that it’s an evil thing to oblige people to vote, well of course it is and nobody does.’[24] Well, nobody other than the Parliament, the Electoral Act, the Australian Electoral Commission, the courts and ultimately the criminal justice system. Lisa Hill of the University of Adelaide is a honourable exception. She defends compulsory voting (although she defines it as attendance at a polling booth) but concedes her ‘arguments may not be enough for those for whom individual rights are always trumps, or others who rail at being forced to be free’.[25] Yet, even she is concerned ‘that highly publicised cases of prosecuted voting recalcitrants are likely, over time, to bring the institution of compulsory voting into unjustified disrepute.’
Compulsory voting is already disreputable – even the leader of the official opposition says it is ‘an evil thing’. No judge has said it is legal to simply turn up on the day. In each case judges have argued it is compulsory to vote – to mark the ballot, and place it into the ballot box. In each case judges have argued voters must express a preference for particular candidates, even if they do not want to do so. It is difficult to reconcile this coercion with the noble ideals expressed by the AEC in their list of advantages. We have shown a smaller group than the entire electorate expresses the will of the people. The same government, following much the same policies, would have been elected at each of the last four elections. Given that compulsory voting allows political parties to better target marginal seats, it may well reduce pork barrelling, but it hardly requires the government to consider the whole of the electorate. Ultimately, compulsory voting is a form of paternalism. Some voters it seems don’t know that they should vote – they don’t know what is good them. Yet, under compulsory voting they do seem to know what is good for the country.
* School of Economics, Finance and Marketing, RMIT University. ARC Grant DP0449846 (Economic reform and Australian electoral decision making) supports the research in this paper.
[1] Electoral Reform in the UK? May 10, 2005. At http://andrewbartlettonline.blogspot.com/2005/05/ electoral-reform-in-uk.html.
[2] See Louis Massicotte, Andre Blais and Antonie Yoshinaka, 2003, ‘Establishing the Rules of the Game: Election Laws in Democracies’, Toronto: University of Toronto Press. In his entry to the International Encyclopaedia of the Social and Behavioral Science, Simon Jackman adds additional countries, but does not systematically list them.
[3] John Hirst has documented the history of compulsory voting, ‘Compelling citizens to vote is a mockery of democracy’, The Age, November 22, 2002. See also his ‘The distinctiveness of Australian democracy’ available at http://www.aph.gov.au/Senate/pubs/pops/pop42/hirst.pdf
[4] See Padraic McGuinness, 1995 ‘The case against compulsory voting’ on this point. Available at http://www.mind-trek.com/writ-dtf/votehoax/p-mcguin.htm.
[5] Email communication from AEC Media & Communications Section, Central Office Canberra.
[6] See for example, the Letters pages of the Australian, or the Sydney Morning Herald on Thursday 6 October 2005.
[7] AEC, ‘Second submission to 2004 Federal Election Inquiry’, pg. 32.
[8] Colin A. Hughes, 1968, Compulsory voting. In, Colin Hughes (ed), ‘Readings in Australian Government’, St. Lucia: University of Queensland Press.
[9] Chris Puplick, 1995, ‘The case for compulsory voting’, Elections Today, 5: 3 – 5. Electronic copy, with commentary, available at http://www.mind-trek.com/writ-dtf/votehoax/c-puplic.htm.
[10] Australian Electoral Commission, 2004, ‘Informal Voting’ Electoral backgrounder No. 18
[11] Rod Medew, 2003, ‘Research Report 1 – Informal Vote Survey House of Representatives – 2001 Election’, Available at www.aec.gov.au (accessed 18/04/2005).
[12] Joint Standing Committee on Electoral Matters (JSCEM), 2005, Inquiry into the Conduct of the 2004 Federal Election and Matters Related Thereto, pg. 183. Available at www.aph.gov.au/house/committee/em/elect04/.
[13] Of course, confusion may also reign due to deliberate efforts to obscure the fact that voting itself is mandatory.
[14] Luke Andrews, Tim Fry and Keith Jakee, 2005, ‘Compulsory voting: Some analytics and a test from Australia’, RMIT mimeo.
[15] Andrew Leigh, 2005, ‘Program paved the road to re-election’, Australian Financial Review, 10 March.
[16] The AES data is available from the Australian Social Science Data Archive. We wish to state that those who carried out the original analysis and collection of the data bear no responsibility for our analysis or interpretation of the data.
[17] S. Jackman, 1999, ‘Non-compulsory voting in Australia?: what surveys can (and can’t) tell us, Electoral Studies 18: 29 – 48.
[18] We thank Andrew Leigh for suggesting Senate inclusion.
[19] In addition to Jackman, ibid, see M. Mackerra and I. McAlister, 1999, ‘Compulsory voting, party stability and electoral advantage in Australia’, Electoral Studies, 18: 217 – 233.
[20] Arend Lijphart, 1997, ‘Unequal participation: Democracy’s unresolved dilemma’, American Political Science Review, 91: 1 – 14.
[21] We do not report the confidence intervals for the Jackman procedure, or our measures. They are available upon request.
[22] Quoted in JSCEM 2005, pg. 203.
[23] We have not modelled preference flows. It is possible, but we believe unlikely, that preferences could have seen the One Nation senator elected, and Brian Harradine elected.
[24] http://www.alp.org.au/media/1005/dsiloo040.php
[25] Lisa Hill, 2002, ‘Compulsory voting: Residual problems and potential solutions’, Australian Journal of Political Science, 37: 437 – 455. See also her 2002 paper ‘On the reasonableness of compelling citizens to ‘vote’: the Australian case’, Political Studies, 50: 80 – 101.
