BETTING ON ELECTIONS: HISTORY, LAW AND POLICY Graeme Orr *

ABSTRACT Betting on elections has a long history, despite periods in which wagers were unenforceable and even criminalised. In recent years, significant online markets have emerged, driven by the bookmaking industry in those jurisdictions which license betting on politics. These markets treat election wagers as a form of sports betting. This article examines the provenance and regulation of election betting in the common law. It charts this from early case law holding wagers involving electors to be void (as tainting voting decisions) through criminal prohibitions, some of which are still on the statute books (since wagers could disguise electoral bribes) and onto contemporary regimes for licensing electoral bookmaking. Normative arguments about election betting and the law include the liberal harm principle, the precautionary principle and the concept of commodification. The article concludes that friendly wagers should be permitted, to allow partisans to intensify the ritual experience of elections. But bets involving politicians should be outlawed, and the industrialisation of election betting should not be encouraged given the risk of commodifying the values underlying electoral democracy.

I

OVERVIEW

Wagering on electoral politics is a burgeoning and special category of gambling. Today, election betting sits alongside sports betting and so-called entertainment and novelty betting, 1 distinct from traditional forms of regulated betting such as horse racing. 2 Some

*

1

Professor, Law School, University of Queensland. This article draws on concepts developed in an ARC project on the law of deliberative democracy (DP130100706). Thanks to Julie Oates and Tim Brown for historical referencing, Greg Dale for research assistance, and Murray Goot, Ron Levy, Warren Swain and Mel Keenan for comments. Appreciation also to NYU Law School (where the article was completed during a Hauser Global Senior Fellowship). Any errors are mine. Entertainment bets are laid on popular culture or artistic affairs (everything from an aspect of a celebrity’s life to the winner of a literary prize). Novelty bets are laid on unusual topics — for examples recently offered, including a book on the punctuality of Melbourne trains, see Tom Cowie, ‘All Aboard for Train Betting, as Novelty Wages Hit the Fast Track’, Crikey (online), 13 June 2012 .

310 Federal Law Review Volume 42 ______________________________________________________________________________________

betting agencies even list election odds as if they were a sub-set of betting on sporting encounters. 3 Agencies run books, offering odds on events ranging from the date a Prime Minister might set for the next election or who will lead the major parties at that election, through to which party will win individual seats. The most popular betting options are the broadest and simplest, namely who will be the next Prime Minister or President, or which party will form government and the size of its majority. 4 Yet election betting is anything but a novel activity. It has a significant heritage, despite it having been formally banned for much of the 20th century in Australia. A criminal prohibition even persists, albeit unenforced, for New South Wales (NSW) elections. As we will see, such censorious legislation did not reflect an explicit wowserism. Rather, it arose from fears of the corrupting potential of widespread election betting on electoral choice, including a fear that private bets could mask electoral bribery. The emergence of the phenomenon of modern election betting has not been driven by private wagering, nor has it been tainted by explicit electoral corruption. Rather, it is the product of a sophisticated sports-bookmaking industry, enlivened by new technologies. Despite this growth, there is a decided regulatory ambivalence about election betting markets. Most Australian States have not approved elections as permissible betting contingencies for licensed bookmaking. Nor have the various States of the US. But a few jurisdictions, such as the Northern Territory, the UK and Ireland are happy to profit from bets made by residents of others. Thanks to the internet, the genie is now out of the bottle. 5 Aside from speculating on whether betting markets are better predictors of election results than opinion polls, academic work is yet to catch up with these developments. This article seeks to fill that void, by examining the legal dimension of election betting across both doctrinal and normative domains. By way of background, and after this introduction, the second section of the article paints a brief history of election betting and places the modern market in financial context. The article then explains the common law of wagering, and unearths the prohibitionist approach to election betting that emerged in Australia from the late 19th century. The fourth section of the article examines the regulation of the modern bookmaking industry, and the complex and fuzzy law around election betting in Australia. The article concludes, in its fifth section, with an assessment of different rationales for the regulation of betting at contemporary elections. It does so in the context of contemporary concerns about the proliferation of gambling in general, and its potentially corrosive effects on those events, like sporting contests, upon which it is parasitic. As we 2 3 4 5

I will use the term ‘election betting’, although betting can extend more broadly to cover events over which electors have no direct influence, like the timing of elections or the fate of legislation. See, eg, Unibet Australia, Betting , listing ‘Politics’ as a category of ‘Sports’. See the Centrebet agency’s account of its election markets: Centrebet, Australian Federal Election Betting — The Basics . Compare sports betting in the US. Whilst underground betting occurred privately, that market only ‘substantially increased’ with the advent of the internet, thanks to sports betting being legal in Nevada: William N Thompson and Tim Otteman, ‘Sports Betting’ in William N Thompson (ed), The International Encyclopedia of Gambling (ABC-CLIO, 2010) vol 1, 206, 209.

2014 Betting on Elections 311 _____________________________________________________________________________________

will see, there is nothing directly corrupting in present day election betting practices — provided politicians do not wager on events in which they are involved. Nor, despite the social harms associated with proliferating opportunities to gamble, does a niche like election betting represent a potent extra risk of such harm. However, as the election betting market grows, we should be concerned about a different form of corrosion, in the form of a commodification and further erosion of the language and values through which we understand politics. It is one thing to respect the freedom of those who are interested in politics to lay friendly wagers with each other, to heighten their psychological enjoyment of electoral politics. It is another thing to argue that an industry of electoral bookmaking, of potentially unlimited scope, should be encouraged by law. Seen in this light, the States’ reluctance to condone election bookmaking shows prudence. But in practice, as long as election betting continues unabated via the Northern Territory, State-based regulation will be largely symbolic and expressive.

II ELECTION BETTING IN CONTEXT Election betting forms a growing source of gambling in western countries generally. There are no firm estimates of turnover in Australia, but it is far from insignificant in its own right. Bookmakers have said the bets they hold on a national election exceed the bets they hold on the final of any particular football code. 6 In 2007, the federal election was the third largest betting event, beside the Melbourne Cup horse-race and the Australian Open tennis. 7 The primary players in framing election betting markets are bookmaking agencies licensed in the Northern Territory: Sportingbet, 8 Centrebet, 9 Sportsbet, 10 and Unibet.11 There is international involvement and capital in the industry. Sportsbet, which grew up in the Northern Territory, was acquired by the Irish betting company Paddy Power in 2010. Unibet, which also operates under the Australian brand Betchoice, is part of an organisation predominantly focused on Europe. A fifth significant player in the Australian election betting scene is Betfair. Though its Australian operations are jointly owned by Victorian based Crown Casino, it is part of a UK group. 12 It differs from the other agencies in operating a betting exchange, under a Tasmanian licence granted in 2006. A betting exchange, like a stock exchange, does not lay odds (prices) or enter bets (transact shares) 6 7

8 9 10 11 12

For similar suggestions from the UK, see Rosenbaum, below n 40, 9. The growth is clear. In 2007 $7 million was said to have been wagered on the national election; in 2010 at least $10 million: Peter Hartcher, ‘Either Way, It’s History in the Making’, The Sydney Morning Herald (Sydney), 24 November 2007, 7; Geoffrey Rogow, ‘Australia’s Election Uncertainty Prompts Betting Bonanza’, The Wall Street Journal (online), 5 September 2013 . In 2013 an offshore betting exchange, Betfair, alone held over $5 million on the outcome of the Australian election. Sportingbet, About Us . Centrebet, About Us . Sportsbet, About Us — Online Betting . Unibet, Welcome to Unibet! . Betfair, About Us .

312 Federal Law Review Volume 42 ______________________________________________________________________________________

itself. So it is not a party to any wagers, but facilitates them for a commission. 13 Betfair arranged bets of over UK£2.5 million on which major party would form government following the 2013 Australian election. The election betting pool, of course, pales into insignificance when compared to the totality of the sports betting market. The election betting share of the gambling industry has been described by the Productivity Commission as one of the ‘[m]inor forms of wagering’ and a ‘very small market’, 14 yet such terms are relative. They are relative to the ‘price’ of electoral democracy to governmental budgets: it cost just $108 million to run the 2010 national election, with an additional $53 million of public funding to defray parties’ campaign costs. 15 They are also relative to the massive size of the overall gambling market. The Productivity Commission valued the sports betting market, in 2008-09, at about $200 million (that is net expenditure by punters). 16 In turn, sports betting forms a fraction of total Australian gambling losses of $19 billion per annum. 17 Gambling is a very significant industry, including in terms of public revenue raised through taxes and licence fees. An important element in the expansion of gambling has been a slow relaxation of cultural factors that had tended to corral gambling to horse racing and the formerly government owned TABs. 18 The other key element has been coincidental developments in technology. The cultural changes include an increasingly secular and market driven approach to society, which accentuates and even valorises that which is quantifiable and tradeable. The cultural story, however, is not a simple one of a shift to legal or social laissez-faire. Whilst the gaming industry has been prominent in resisting regulation in the past decade, 19 so too have anti-gambling groups such as the Australian Churches Gambling Taskforce. This backlash is reflected in the repeat electoral successes of independent politicians running on anti-gambling platforms, like Senator Xenophon and Andrew Wilkie MHR. (Both MPs have called for election betting to be outlawed.) 20 Technological changes conducive to gambling include the emergence of the internet as a rapid and cost effective form of communicating odds, offering betting options, and transacting wagers and payouts. The High Court has taken judicial notice of the 13 14 15 16 17 18 19 20

See Betfair Pty Ltd v Western Australia (2008) 234 CLR 418, 466 [56]–[60]. This allows punters to frame their own odds and options, including to lay a bet, i.e. bet against a particular outcome. In 2013 Betfair also commenced a ‘Sportsbook’, through which it lays odds directly. Productivity Commission, Gambling, Report No 50 (2010) vol 1, 2.37. Australian Electoral Commission, Electoral Pocketbook (Australian Government Publishing Service, 2011) 77. Productivity Commission, above n 14, figure 2.1 at 2.5. Total betting revenue is much higher, but the money ‘churns’, as winning bets are sometimes ‘reinvested’ in another bet. Ibid. The social dimension of the problem is captured in the calculation that betting losses represent 3.1% of average household expenditure, larger than the net expenditure on alcohol. ‘TAB’ stood for Totalisator Administration Board, a bureaucratic name for a mechanical system of bookmaking. Graeme Orr and Anika Gauja, ‘Third Party Campaigning and Issue Advertising’ (2014) 60 Australian Journal of Politics & History 73. Australian Associated Press, ‘Independent MPs Want Election Betting Ban’, The Australian (online), 31 January 2013 .

2014 Betting on Elections 313 _____________________________________________________________________________________

technological shift, as well as the jurisdictional challenges it poses to an area of law traditionally left to individual States and Territories: 21 There is a developed market for the provision throughout Australia of wagering services in respect of horse racing and other sporting events. Those events may take place in one State, the customer may be located in another State and the provider of the wagering services may do so from a third State. This geographic separation is reduced not only by telephony but also by the omnipresence of the internet and the ease of its use. 22

The heritage of election betting Whilst the burgeoning, formalised election betting market run via licensed bookmakers and betting exchanges is a contemporary phenomenon, betting on elections and indeed broader political events (such as the outcome of military adventures) is far from a modern invention. Baumgartner, a historian of the Vatican’s unique electoral politics and systems, asserts that betting on papal elections was well established before the 16th century. 23 An account of Pennsylvanian elections in the early 1800s records a culture of significant public betting around local polling, augmented by drinking. It was so rife that a censorious visiting Englishman was moved to observe that ‘[t]hey were all betting upon the election; but I lament to say, that few, if any, appeared to care one straw about principle.’ 24 Betting generally — and concerns for its restriction — is of course even older. In the common law world, legislation to suppress betting via games of chance dates back at least to the Unlawful Games Act of 1541–2. 25 In their recent survey of international experience, Rhode and Strumpf affirm that betting on politics has a long and broad history, at least in the West. 26 The focus, especially since the rise of modern electoral democracy in the early twentieth century, 27 has been on the outcomes of major elections, like the US Presidency and parliamentary majorities in the UK. They reason that ‘pivotal elections energized [the] market’. 28 In other words, competitive electoral events that also involve a sense of historical import or social contest, periodically fuel markets in political betting. In a separate paper, Rhode and Strumpf show that betting on US Presidential elections was rife at the turn of and into the 21

22 23 24 25

26 27 28

The Commonwealth nonetheless can use its power over trading corporations under the Australian Constitution s 51(xx), to regulate the gambling industry: See, eg, National Gambling Reform Act 2012 (Cth) ss 55, 58, 60, 79 (gaming machine premises and manufacturers). It has also used its telecommunications power under the Australian Constitution s 51(v) to prohibit internet casinos whilst preserving online wagering: Interactive Gambling Act 2001 (Cth) ss 6, 8A. Betfair Pty Ltd v Racing New South Wales (2012) 249 CLR 217, 254 [2] (joint judgment). Frederic J Baumgartner, Behind Locked Doors: A History of Papal Elections (Palgrave Macmillan, 2003) 88. Mark W Brewin, Celebrating Democracy: The Mass-Mediated Ritual of Election Day (Peter Lang Publishing, 2008) 82. An Act for Maintenance of Artillery and Debarring of Unlawful Games, 33 Henry 8, c 9. Section 8 forbade for profit gaming houses, including the games of ‘Dysing, Table or Cardinge’ [dice, backgammon or cards]. Yet ‘gentlemen’ and ‘noblemen’ were permitted to play with their servants and families (ss 15–16). Paul W Rhode and Koleman Strumpf, ‘The Long History of Political Betting Markets: An International Perspective’ in Leighton Vaughan Williams and Donald S Siegel (eds), The Oxford Handbook of the Economics of Gambling (OUP, 2013) 560. Demarked by a universal franchise and settled parties. Rhode and Strumpf, above n 26, 560.

314 Federal Law Review Volume 42 ______________________________________________________________________________________

early twentieth century. 29 They argue that, despite attempts to manipulate odds, favourites usually won, even at a time before scientific opinion polling (which dates to Gallup in 1935). Yet election betting declined in the mid-twentieth century, as polling rose and other forms of betting proliferated. It has re-emerged in the US in the electronic era, albeit in a legally truncated way. American state law either bans political betting, or does not authorise bookmakers to take bets on such events. 30 The Federal Wire Act of 1961 criminalises internet gambling, if only on a ‘sporting event or contest’. 31 The Unlawful Internet Gambling Enforcement Act of 2006 further limits the market. 32 Nonetheless, two outlets have emerged in the US. One is via overseas markets, the most prominent being the UK-based Betfair, and the Irish ‘prediction market’ Intrade. 33 Intrade alone transacted ‘contracts’ worth almost US$100 million on the 2012 Presidential election, 34 before encountering unrelated legal problems and insolvency. The other is the Iowa Electronic Market, an academically-run futures market. Although operating in a grey legal area, it has been spared prosecution provided it remains a not-for-profit. 35 Electoral wagering in Britain also has a significant lineage. As part of an historical study, Beers notes records of personal bets on politics in Oxbridge ‘college betting books’ from the late 18th century. 36 Her study focuses, however, not on such one-to-one wagers, but on uncovering the history of ‘[m]ajorities trading’ by stockbrokers between the world wars. Post-dating similar developments in the US, such trading was ‘essentially [a] futures market in party stocks’. 37 That is, it was speculation on the likely seat numbers each party 29 30

31 32 33 34

35 36 37

Paul W Rhode and Koleman Strumpf, ‘Historical Presidential Betting Markets’ (2004) 18 Journal of Economic Perspectives 127, 127–42. This is based on calculations from contemporary newspaper sources. Even in Nevada, the home of gambling, betting on elections is a ‘gross misdemeanour’: Nev Rev Stat 2011 §293.830. A bill to repeal this was recently rebuffed: Sean J Miller, ‘Nevada Passes on Federal Bet Law’, Campaigns and Elections (online), 13 June 2013 . See 18 USC §1084; In re Mastercard International Inc. Internet Gambling Litigation, 313 F 3d 257 (5th Cir, 2002). See 31 USC §5361; Andrew S Goldberg, ‘Political Prediction Markets: A Better Way to Conduct Campaigns and Run Government’ (2010) 8 Cardozo Public Law, Policy and Ethics Journal 421, 431–8. A prediction market is an exchange for speculating (in effect gambling) on future events. David M Rothschild and Rajiv Sethi, Trading Strategies and Market Microstructure: Evidence from a Prediction Market (8 September 2013) Social Science Research Network < http://ssrn.com/abstract=2322420> (an econometric study showing that 1.2 million US$10 contracts were made in the final fortnight of that campaign). John Ydstie, They Call the Election a Horse Race; It has Real Bettors, Too (19 October 2012) NPR (citing US$85 million traded before that fortnight). Melonyce McAfee, ‘Are “Political Futures” Illegal?’, Slate (online), 27 April 2007 . Laura D Beers, ‘Punting on the Thames: Election betting in Interwar Britain’ (2010) 45 Journal of Contemporary History 282, 282. Ibid 286.

2014 Betting on Elections 315 _____________________________________________________________________________________

would win at a forthcoming election. (We would now call this ‘spread betting’. The main difference is that contemporary election betting is via bookmakers rather than stockbrokers, and is more open to ordinary punters and less centred on the rentier class).38 Majorities trading grew rapidly from an informal system of wagering between brokers, into a high-volume business. Its development, prior to opinion polling, attracted great interest from a political elite and press searching for predictions or reassurances about election results. However the UK market proved a failure in predictive terms. 39 In contrast, contemporary scholars from economistic backgrounds promote the predictive value of diversified election betting pools. 40 Election betting today often follows published opinion polling, but may also be informed by ‘insider’ information, 41 or distorted by ideologically motivated bets.42 Regardless of its predictive value or foibles, election betting re-emerged in the UK in the 1960s, with the 1966 election being the largest betting event in UK history at the time; and it has blossomed further with the advent of electronic betting. 43 Betfair’s 2011 annual report named the UK general election of 2010 as the top ‘special’ betting market of that year. 44 Election wagering in Australia was imported as a cultural artefact from Britain, and as we shall shortly see, spurred a variety of prohibitions in Australian electoral acts. Presumably as a result of those prohibitions, election betting was modest in Australia until recently, although newspapers occasionally reported on bookmakers’ odds and the size of individual bets. 45 The first of the contemporary agencies to run a public book on Australian elections was Sportsbet, on the 1987 Northern Territory and national

38 39 40

41 42

43 44

45

The language is instructive: commentators from economic backgrounds like Rhode and Strumpf, above n 26 prefer terms like ‘markets’ and ‘traders’, as if election betting were a futures market rather than a form of punting. Because the UK market lacked opinion poll data to inform it and, given the limited class of capitalists who drove it, the market was not broad enough to be a real crowd-sourcing exercise. See, eg, Andrew Leigh and Justin Wolfers, ‘Competing Approaches to Forecasting Elections: Economic Models, Opinion Polling and Prediction Markets’ (Working Paper No 1972, Institute for the Study of Labor, 2006) 325. More equivocally, see Martin Rosenbaum, ‘Betting and the 1997 British General Election’ (1999) 19 Politics 9. See generally Erik Snowberg, Justin Wolfers and Eric Zitzewitz, ‘Prediction Markets for Economic Forecasting’ in Graham Elliott and Allan Timmermann (eds), Handbook of Economic Forecasting (Elsevier, 2013) vol 2, 657. Examples include: high-quality polling known only to party insiders, advanced information about preference deals and candidate withdrawals or major policy announcements. Leigh and Wolfers, above n 40, suggest betting markets provide better predictions at the individual seat level. They also claim betting markets are less volatile than polls conducted well-prior to an election. But this misses two points. First, opinion polls only purport to be a snapshot of the sentiment at a particular time. Second, polling is a democratic feedback mechanism: politicians adjust their positions to mollify popular opinion. Rosenbaum, above n 40, 9. ‘Special’ here means novelty or entertainment markets. Markets on a reality television show and the Eurovision Song Contest were the next largest: Betfair Group PLC, Annual Reports and Accounts 2011, Betfair, 23 < http://corporate.betfair.com/~/media/Files/B/BetfairCorporate/pdf/annual-Report-2011.pdf>. Rhode and Strumpf, above n 26.

316 Federal Law Review Volume 42 ______________________________________________________________________________________

elections. 46 As we noted, the market’s recent flourishing is a product of evolving technology and culture.

III HISTORICAL REGULATION OF ELECTION BETTING: THE COMMON LAW AND AUSTRALIAN LEGISLATION What was the attitude of the common law to election wagers? In particular, were they enforceable? This is a matter of contract law, dependent on whether wagering on elections was judicially deemed to be contrary to public policy. On this question — indeed on the broader question of the desirability of wagering generally — the common law fluctuated. Wagers per se were lawful and hence were enforceable at common law; yet this permissive approach was subject to numerous judge-made exceptions. (Such judicial policy could be influenced by statute or divined from sensitivities about public morality; including the potential that betting would infect the event bet upon.) A House of Commons report of 1844 summarised the law on wagering by observing that judges, from Georgian times, had become more interventionist and censorious about wagers whose subject matter they considered inappropriate. 47 Bentham, similarly, had come to see wagering as suspicious, potentially ‘subornation for all imaginable crimes’: 'Bet, for example, that a certain person does not live beyond a certain time, and trace the consequences which such a bet may have.' 48 What of political betting particularly? In Stuart times, a wager that Charles II would return from exile to reclaim the throne within a year was upheld — despite such wagers potentially corrupting the highest affairs of state. 49 Yet by 1785, in Allen v Hearn, 50 a bet over an election for a parliamentary constituency was held to be unlawful. This was despite the bet involving partisan supporters of rival candidates. In that case, Allen was seeking £100 winnings from the defendant. Allen argued that the wager involved no risk to the public interest. But Lord Mansfield CJ reasoned that however playful this particular wager, betting on elections risked the freedom and purity of elections, at least where an elector was a party to the wager. This was because such bets introduced a pecuniary element into voting decisions. 51 The case, it should be noted, was decided prior to the Reform Act, when voting rolls were relatively small and individual votes had more determinative influence. The mere fact that the common law rendered election wagers unenforceable was probably of limited deterrence to election betting and any consequent pollution of voting decisions. Settling bets has always involved honour between friends, or the reputation of a bookmaker. Candidates also came to realise that by wagering against themselves, they 46 47 48 49 50

51

David Nason, ‘Raising the Stakes’, The Australian (Sydney), 1 October 1998, 13. Report from the Select Committee on Gaming (1844) Appendix I (compiled by T Starkie QC). Jeremy Bentham, ‘General View of a Compete Code of Laws, Ch XVI’ in John Bowring (ed), The Works of Jeremy Bentham Vol 3 (William Tait, 1843) 1070. Andrews v Herne (c 1673) 1 Lev 33; 83 ER 283. (1785) 1 TR 56; 99 ER 969. Compare Jones v Randall (1774) 1 Cowp 37; 98 ER 954, upholding a wager about the fate of a court appeal as a kind of litigation insurance, provided the wager did not involve the lawyers or judges in the case. For discussion of the seminal 18th-19th century decisions on the legal status of wagers, see Warren Swain, ‘Da Costa v Jones’ in Charles Mitchell and Paul Mitchell, Landmark Cases in the Law of Contract (Hart, 2008) 119, 127–34. Allen v Hearn (1785) 99 ER 960, 971.

2014 Betting on Elections 317 _____________________________________________________________________________________

could mask briberous inducements to individual electors, without even requiring a stake from the elector. When the UK Parliament came to enact the ground-breaking Corrupt and Illegal Practices Prevention Act 1883, however, it did not include a specific criminal offence of wagering on elections. (Wagering had been declared void, but not criminal, by an Act of 1845.) 52 The 1883 Act instead focused on expanding the war against specific corruptions to the general challenge of restricting the cost of electioneering. Strict limits were thus placed on how much a candidate could spend, and all expenditure had to be channelled via election agents. To borrow modern economic jargon, the legal focus shifted from microeconomics (prohibited transactions) to macroeconomics and the ‘supply-side’ of the question of money and elections. 53 Australian prohibitions Election betting attracted legislative attention in Australia as early as 1856, when the Victorian Legislative Council made it an offence, subject to a £20 fine, ‘to make any wager bet or other risk’ on any election to the Council or the new Legislative Assembly. 54 Unfortunately, there are neither Hansard records of the debates, nor any clue in the newspapers of the time as to what motivated this novel provision. The great interest of the day was inaugurating the secret ballot and expanding the franchise. The ban on election betting fell within an Act specifically dealing with electoral purity and corruption. 55 The Legislative Council had, to that time, been elected on a narrow property and educational franchise. 56 But whether banning election betting was a prophylactic prompted by problems with wagering on elections in British practice, or simply generated by distaste that something as precious as an emerging democracy would be the subject of betting, is not knowable. The Victorians may have been aware of prohibitions in parts of the US. 57 In NSW, a similar offence and fine was erected in 1893, 58 and the legislative debate is available. The predominant reason given in the NSW Legislative Assembly was concern that wagering often masked electoral bribery. 59 However if that were the only rationale, the ban could have been directed purely at wagers involving candidates or their agents, especially those who laid odds against themselves.

52

53 54 55 56 57 58 59

Gaming Act 1845, 8 & 9 Vict c 109, s 18. Betting on horse racing was exempt. This rule remained until 2007. A person could only sue to recover a bet and then only if they called it off prior to the event wagered upon: George Henry Hewitt Oliphant, The Law Concerning Horses, Racing, Wagers and Gaming (Sweet, 1847) 204. Graeme Orr, ‘Suppressing Vote-Buying: the ‘War’ on Electoral Bribery from 1868’ (2006) 27 Journal of Legal History 289. Election Proceedings Regulation Act 1856 (Vic) s 28. Anon, ‘New Laws: Elections Regulations Act’, The Argus (Melbourne), 15 April 1856, 4. Gerard Carney, The Constitutional Systems of the Australian States and Territories (Cambridge University Press, 2006) 54. Brewin, above n 24, 268–9 notes two prohibitions legislated in Pennsylvania in the 1810s and 1830s (in Brewin’s endnote 74). Parliamentary Electorates and Elections Act 1893 (NSW) s 120. New South Wales, Parliamentary Debates, Legislative Assembly, 10 November 1892, 1776–7.

318 Federal Law Review Volume 42 ______________________________________________________________________________________

In debate in the Legislative Council, in-principle opposition to election betting was evident. Indeed, many members wanted harsher penalties and even gaol time. Their reasons were a mix of antipathy to gambling generally, and concern that speculation on elections was corrupting. 60 Corruption included disguised bribes, but it extended to a broader sense of corrosion of an electoral ideal. This embraced concern that those with a financial stake in an election had an incentive to swing it improperly, 61 through to a fear that bookmakers might frame odds to influence the votes of electors who bet with them. 62 There was also a more general sense that betting tainted elections by ‘associations of an immoral character’ and that it was a ‘disgrace to the country that elections should be made the subject of gambling’. 63 Only one member suggested the risk was overstated. 64 The clear consensus was that all election betting should be rendered illegal, however innocent the intention. 65 In the first national electoral law of 1902, wagering on Commonwealth elections was similarly criminalised, with a maximum penalty of £50. 66 This provoked more than a modicum of parliamentary debate. The predominant fear was of widespread wagering as a cover for corruption. One senator alleged that a Tasmanian Minister of the Crown who had been unseated for illegal practices nonetheless stood again for election, and that his supporters ‘went through the electorate and said to elector after elector —“I will bet you so much that [the former minister] does not get in.” The result was that the people with whom the wagers were made voted for the candidate, and he was again returned.’ 67 This appears to be a reference to Brighton Election Petition; Mugliston v Dillon. 68 The Tasmanian Supreme Court there unseated Dillon, whose sins of bribery — and treating — 69 included two counts of briberous wagering. Evidence of one elector captures the flavour: This was about a month or six weeks before the election. Dillon and I were the only two people in the dining room [of a hotel]. Dillon asked me what I thought of the election, what sort of chance he had. I told him I thought he would be returned. He offered to make a bet of £5 to £1 that he was not. I said I will take the £5 that you are returned. We shook hands … There was another bet a short time before the election … We talked the election over, and I

60 61 62 63 64 65 66 67 68 69

New South Wales, Parliamentary Debates, Legislative Council, 21 February 1893, 4374–9 (Alexander Brown, Richard Edward O’Connor, William Montagu Manning and Leopold Fane de Salis). Ibid 4376 (Alexander Brown) and 4377 (John Smith). Ibid 4377 (William Montagu Manning). Ibid 4378 (Alexander Brown and William Montagu Manning respectively). Ibid 4375 (Henry Carey Dangar). Contrast allegations that hundreds of pounds were bet in a single electorate: 4376 (Alexander Brown). See, eg, ibid 4377 (Richard Edward O’Connor) and 4378 (John Mildred Creed). Commonwealth Electoral Act 1902 (Cth) s 182. Commonwealth, Parliamentary Debates, Senate, 13 March 1902, 10949 (John Barrett). (1891) 13 ALT 44 (‘Brighton Election Petition’). For similar allegations involving a former New Zealand Premier, see Anon, ‘The Stout Election Petition’, The Timaru Herald (New Zealand), 4 January 1894, 3. The petition was dismissed on unrelated technical grounds. Treating was an old-fashioned form of vote buying through lavish alcohol and food.

2014 Betting on Elections 319 _____________________________________________________________________________________ bet Dillon a level sovereign that he headed the poll. … After the £1 bet I bet him a champagne supper that he would be returned. Mr Dillon made the offer of that bet. 70

Dillon admitted the conversations; the elector admitted he voted for Dillon. The Court had no qualms finding the allegations proven, given other stark evidence in which Dillon, on being rebuffed by a voter, had said ‘If I give you three guineas will you vote for me, and induce other members of your cricket club to vote for me?’ 71 The amounts on offer, especially in the bets, are noteworthy. Princely sums were not unusual in a betting culture, nor necessarily uneconomic where the electorate was relatively small and voting was still voluntary. Revealingly, the elector said he had been dissuaded from collecting his winnings, having been told Dillon would never pay up. Briberous wagering was less expensive than it first appears: unless a trusted stakeholder was used, short of blackmailing the candidate the wager was legally and practically unenforceable. One federal member wanted electoral wagering of any sort to be made an offence equal to bribery, a suggestion the Government did not treat seriously enough for his liking: The Minister for Trade and Customs may smile, but he would not do so had he had my experience of the results of wagering in elections. It is extremely dangerous to allow people to wager in this way. 72

He was supported by a fellow Western Australian who claimed to know of candidates who had ‘laid wagers with individuals who were able to influence a large number of voters, that they would not win’, a practice that was ‘one of the most insidious and objectionable methods of bribery.’ 73 On the other side stood parliamentarians who felt that the general proscription of bribery was sufficient for candidates who laid bets against themselves. One Senator felt that outlawing all betting on elections was tantamount to cultural insensitivity, since ‘in Western Australia it is the invariable custom for nearly every one to bet on the result of an election’. 74 Unlike in the NSW Legislative Council a decade prior, wowserism was not a noticeable legislative motivation. The precautionary principle in relation to betting’s corrupting effects on electoral behaviour was the prominent reason. As Sir Edward Braddon, the oldest MHR of the time put it: I have known instances of candidates backing themselves simply as a sporting matter. However, I am willing that persons should be prevented from entering into wagers even as harmless as that. 75

The Leader of the Government in the Senate, Richard O’Connor — who had been a prominent proponent of the NSW reform — also noted a gradation of motives behind

70 71 72 73 74 75

Brighton Election Petition (1891) 13 ALT 44, 45–6. It seems odd that a candidate would keep betting with the same elector; but he may have wanted to keep in the elector’s good books, and each bet represented insurance. Ibid 46. He was still rebuffed. Commonwealth, Parliamentary Debates, House of Representatives, 24 July 1902, 14647 (Hugh Mahon). Ibid (James Mackinnon Fowler). Commonwealth, Parliamentary Debates, Senate, 13 March 1902, 10949 (Alexander Perceval Matheson). Commonwealth, Parliamentary Debates, House of Representatives, 24 July 1902, 14647 (Sir Edward Nicholas Coventry Braddon) (emphasis added).

320 Federal Law Review Volume 42 ______________________________________________________________________________________

betting. He did not expect innocent private or friendly wagers to attract even a nominal fine of a shilling. 76 The ban on election betting was soon copied into the Western Australian Electoral Act 1907 and the South Australian Electoral Code 1908. 77 It also found its way into the Queensland Criminal Code 1899. 78 A 1977 stocktake of electoral offences showed that all States and the Commonwealth maintained criminal penalties for election betting. But most of those offences were quietly expunged as electoral acts were modernised in recent decades. 79 The Commonwealth offence was omitted in the legislative overhaul of 1983.80 The criminalisation of wagering on Tasmanian elections survived rewrites of the law in 1985, but not 2004. 81 The one holdout is NSW. Wagering on NSW elections has thus formally remained an offence for 120 years. 82

IV CONTEMPORARY REGULATION IN AUSTRALIA A prohibition such as the NSW offence applies, at least in theory, to a resident of the State who places money by phone or internet with an agency outside the State.83 Yet as recently as the 2010 NSW election such outside agencies ran books on NSW seats. 84 It would seem the criminal prohibition is viewed as symbolically important enough to be retained. But — given the limited fine, the cost of investigation and the apparently victimless offence — it has not been seen as worthy of enforcement in any individual case. The problem with this old-fashioned law is that it conceives of betting as a matter of individual morality, rather than a systemic issue. The regulation of the contemporary gaming industry in Australia is complex, but a taste of it is essential to understand the place of election betting in that regime. Early 20th century gambling regulation was driven by concerns about ‘vice’ and nuisance in the physical dimensions of gaming. This focused on curbing aspects of working class gambling, like face-to-face solicitation, betting in public places and gambling dens. Some

76 77 78 79 80 81 82 83

84

Commonwealth, Parliamentary Debates, Senate, 13 March 1902, 10949 (Richard Edward O’Connor). Electoral Act 1907 (WA) s 188; Electoral Code 1908 (SA) s 189. Criminal Code 1899 (Qld) s 110. P D Finn, ‘Electoral Corruption and Malpractice’ (1977) 8 Federal Law Review 194, 229. It had become part of the Table of Offences in Commonwealth Electoral Act 1918 (Cth) s 170, which was overhauled in 1983 following Joint Select Committee on Electoral Reform, Parliament of Australia, First Report – Electoral Reform (1983) ch 13. Electoral Act 1985 (Tas) s 248, with a maximum fine of $500, repealed by Electoral Act 2004 (Tas). Parliamentary Electorates and Elections Act 1912 (NSW) s 154. The potential penalty (currently $330) has decreased in effective value, compared to the original £20 fine. Whether it could apply to people, or wagers made, outside NSW depends on whether the actions had sufficient ‘geographical nexus’ with NSW: Crimes Act 1900 (NSW) s 10C. If election betting is largely a product of whether an election is ‘pivotal’, most State election betting would be by residents of that State. A State can prohibit a type of betting. But if it allows it, Australian Constitution s 92 prevents protectionist measures like prohibitions on interstate operators competing in that market: Betfair Pty Ltd v Western Australia (2008) 234 CLR 418. Alicia Wood, ‘Bookies Bet on Labor Rout’, The Sydney Morning Herald (online), 27 February 2011 .

2014 Betting on Elections 321 _____________________________________________________________________________________

of that recherché regulation survives. 85 But only the Australian Capital Territory and the Northern Territory have specific unlawful betting regimes, making it an offence to bet with a person who is not a licensed bookmaker. 86 (Some non-commercial, private bets are exempt, if they are side bets on games: for example a domestic poker match in the ACT.)87 Oddly then, in those two territories it is illegal for friends to wager on an election. Any threat of prosecution is muted by the requirement that the participant ‘know’ the bet is unlawful. 88 Such apparently censorious regimes are in truth less concerned with cracking down on wagering between individuals than with protecting a cartel of licensed bookmakers. The Northern Territory in particular is liberal in its approach to licensing, so the major online agencies operate under Northern Territory licences, which permit election betting. The first online betting licences were granted by the Northern Territory in 1995; there are now a dozen companies licensed in that jurisdiction. 89 The Northern Territory regulator (its racing commission) estimates those betting agencies have a total turnover of $6bn pa. The Productivity Commission describes online betting as a Northern Territory ‘export’ industry. 90 The picture in most State jurisdictions is different: NSW, Queensland, Western Australia and South Australia have chosen not to license election betting. In NSW, as discussed, there remains a specific criminal prohibition on betting on NSW elections. The Unlawful Gambling Act 1998 (NSW) allows betting on any approved ‘event or contingency’ with a licensed bookmaker, including with a bookmaker licensed in another jurisdiction. 91 However the Minister has not approved betting on any elections. 92 Instead, the Minister and his Opposition counterpart gave undertakings in 2010 ‘that in extending the range of [sports bets] that may be covered, betting on extreme sports and on elections will not be approved activities.’ 93 Hence one cannot gamble on any election with a NSW

85 86 87 88 89

90 91 92 93

Vestigially, in State criminal codes or bespoke legislation: see, eg, Criminal Code 1899 (Qld) s 232 (operating a place for unlawful games). This remains in a chapter of the Code dealing with ‘Nuisances [and] misconduct relating to corpses’. Unlawful Betting Act 1989 (NT) s 4 (bet is void), s 18 (offence by punter), ss 29 and 31 (offence by person soliciting wagers); Unlawful Gambling Act 2009 (ACT) ss 24, 25 and 29 (offences of arranging, conducting or participating in unlawful betting), s 47 (bet is void). Unlawful Gambling Act 2009 (ACT) s 10. Unlawful Gambling Act 2009 (ACT) s 29(2)(b). Department of Business (NT), Racing and Bookmakers (30 June 2014) . Bookmakers are licenced under the Racing and Betting Act 1983 (NT). Productivity Commission, above n 14, 2.3. Unlawful Gambling Act 1998 (NSW) ss 8(6)(d)-(d1). Betting with unlicensed bookmakers in NSW is otherwise illegal, but ‘bookmaker’ implies a commercial operation (s 4). The Minister could do that, at least for betting on federal elections, by making it a ‘declared betting event’ for licensed bookmaking: Racing Administration Act 1998 (NSW) s 18. New South Wales, Parliamentary Debates, Legislative Assembly, 30 November 2010, 28565 (George Souris). I am indebted to Sonja Hewison of the NSW Electoral Commission for this reference.

322 Federal Law Review Volume 42 ______________________________________________________________________________________

bookmaker. 94 Nor can a person in NSW legally either make an informal election wager with a friend, or place a bet with an interstate bookmaker, on a NSW election result. The South Australian regulator has also not approved betting on political or electoral outcomes as an approved betting contingency, 95 and this includes contingencies offered by interstate operators. 96 (Curiously, Academy Award winners are an approved betting event.) Ostensibly, this means interstate operators should not enter wagers with South Australians about their elections. 97 Yet Sportsbet, for example, ran a book specifically on which party would win the 2014 South Australian election. 98 In Queensland, the Minister has power to approve betting contingencies for sports betting licensees. 99 To date only TattsBet has such a licence. 100 But, echoing South Australia (from where that agency also operates), the only Queensland non-sports and racing events on which agencies can offer odds are the Academy Awards. 101 The Western Australian Gaming and Wagering Commission likewise has discretion to approve certain sports-like contingencies on which licensed bookmakers may trade.102 It has not approved election betting. 103 Western Australia specifically permits ‘social’ betting, however, and this would permit private electoral bets. 104 In contrast, the Victorian Commission for Gambling and Liquor Regulation has listed Australian and even UK elections as ‘approved betting events’ for fixed odds bookmakers licensed in Victoria. 105 Under the Gaming Regulation Act 2003, this means that the Victorian Commission considers election betting not to be against the public interest and a 94

95 96 97

98 99 100 101 102 103 104 105

See the complaint of a Tabcorp chief executive that his company could not take over-thecounter bets in NSW, but that residents could place such bets online with, for example, a Northern Territory bookmaker: Rebecca Urban, ‘Tidy up Rules, Urges Tabcorp Boss’, The Australian (online), 17 March 2012 . Independent Gambling Authority of South Australia, Approval of Contingencies ― Major Betting Operations Licence, 03/2010, 11 December 2001. Independent Gambling Authority of South Australia, Approved Contingencies (Authorised Interstate Betting Operators) Notice 2009, 01/2009, February 2009. This is because ‘conducting betting operations’ includes an interstate bookmaker accepting a bet from someone in South Australia, and interstate operators may only take bets from South Australians on licensed races and approved contingencies: Authorised Betting Operations Act 2000 (SA) ss 40A(4)(c) and 40A(6)(a). In contrast, Sportingbet does not run a book on South Australian elections given the legal position: see Australian Associated Press, above n 20. Wagering Act 1998 (Qld) ss 7, 17 and 56. Queensland Government, Office of Liquor and Gaming Regulation, Race or Sports Wagering (23 April 2014) . (TattsBet was formerly known as Unitab.) Joint Select Committee on Gambling Reform, Parliament of Australia, Interactive and Online Gambling and Gambling Advertising (2011) 200. Betting Control Act 1954 (WA) ss 4A and 4B. Gaming and Wagering Commission, ‘Approval for Sports Betting Events and Contingencies’, Government Gazette WA, 31/3/2009, 1031–5. Gaming and Wagering Commission Act 1987 (WA) s 64(3). Victorian Commission for Gambling and Liquor Regulation, Approved Betting Events (27 March 2014) .

2014 Betting on Elections 323 _____________________________________________________________________________________

reasonable extension of the scope of gambling. 106 Informal or private electoral wagering is not illegal (provided it is not done as a quasi-business) but such wagers are void in the sense of unenforceable. 107 The Victorian approach is thus the inverse of the Western Australian. In 2009, the Tasmanian Gaming Commission also listed ‘Political Elections’ in a liberal list of almost 100 ‘approved sports events’. 108 In short, all the States have empowered their own TABs and licensed bookmakers to take bets on sporting encounters. This reflects each State’s desire to protect the finances and brands of those agencies. But outside the Northern Territory, Victoria and Tasmania there has been a conscious regulatory decision to not approve electoral bookmaking. In most jurisdictions (outside NSW elections), private wagers on elections are not illegal. In particular, there is no law prohibiting betting on Commonwealth elections. However whether private wagers are void (Victoria), explicitly sanctioned (Western Australia) or left to the vagaries of the common law is another matter. As we saw in Allen v Hearne, at least where an elector was involved, electoral bets were unenforceable under the common law; but judicial policy waxes and wanes, and a contemporary judge might find the reasoning in that case outmoded. In recent years, regulatory concern, of course, has hardly been focused on the relatively niche question of betting on politics. Rather, much energy has been spent considering: (1) the proliferation of broadcast advertising of gambling and its exposure to children; 109 and (2) heavy punters or ‘problem gamblers’, and the propensity for casino and electronic gaming to be designed to trigger their addictions. The Senate’s 2010 reference to the Joint Select Committee on Gambling Reform did embrace ‘the implications of betting on political events, particularly electoral outcomes’. However, the Committee made no comment on the issue, aside from noting two competing arguments. One, from an anti-gambling lobby, echoed Allen v Hearn’s old concern that an elector who punted on an election might feel compelled to vote that way. In response, an industry lobby argued that elections were so publicly scrutinised that there should be little concern for the integrity of elections or the integrity of electoral betting. 110 In contrast, betting scandals have undermined faith in sports such as cricket, where the number of matches has spiralled, rendering each encounter less meaningful. Syndicates have corrupted some players with bribes to throw matches and to fix exotic betting ‘events’ like spot and negative options.

106 107 108

109 110

Gambling Regulation Act 2003 (Vic) s 4.5.8. Ibid s 2.4.12. Tasmanian Gaming Commission, ‘Approved Sports Events’ (August 2009) . The power lies under Gaming Control Act 1993 (Tas) s 3(8). Other approved, non-sports categories are ‘Celebrity Announcement’, ‘Cultural Event’, ‘Entertainment – Media’ and ‘Statistics Event’ (sic). Private election wagering is lawful, provided it is not done in a public place (s 5A(2)) or as a business (ss 5A(1) and 76B). Joint Select Committee on Gambling Reform, above n 101; Joint Select Committee on Gambling Reform, Parliament of Australia, The Advertising and Promotion of Gambling Services in Sport (2013). Joint Select Committee on Gambling Reform, above n 101, 157–8.

324 Federal Law Review Volume 42 ______________________________________________________________________________________

It should be remembered that betting agencies have incentives to run books on elections beyond the obvious attraction of increasing revenue and profits. Betting agencies piggyback on the high profile of politics and elections in the media. 111 Their brands and reputations benefit from discussion of the odds they offer, especially on key questions such as the survival of leaders and formation of governments. Although the business model is different, there is an analogy with market research companies, which conduct political opinion polling as a ‘loss leader’ since reporting of their polls boosts their commercial brand.

V ELECTION BETTING: A NORMATIVE ANALYSIS As long as elections are taken seriously as political events and are not subsidiary to gambling on them, then their formal legitimacy is not threatened. Elections are not mere games of chance that can be multiplied endlessly to generate opportunities for betting. Nor are they identical to professional sporting encounters, where the skilled element is easily overwhelmed by a financial desire to manipulate odds. 112 Political power, after all, is its own motivation, and elections are not daily occurrences like horse races or sports events. Yet for the gambling industry only to be concerned with the internal integrity of elections or their own books misses a broader cultural question about the nature of politics. Political betting is not entirely divorced from sports betting. Each is parasitic on a professional enterprise — politics, or a game — that exists prior to any wagering and for autonomous purposes. In the case of politics, there is the pursuit of power and ascertainment of a common good. In sports, there is a competitive game according to confined rules. (In contrast, the bulk of gambling occurs on events artificially organised for no other purpose: horse-racing, lotteries, casinos and gaming machines.) In one conception, elections are also games, if especially serious ones. Talk about elections is thus laced with competitive sporting and even racing metaphors. Thus we speak, in the ‘horse race narrative’, of those ahead in opinion-polls as ‘front-runners’, and of the simplest voting system as ‘first-past-the-post’. 113 In what follows, I want to examine three possible rationales for regulating election betting. One is the relatively narrow liberal harm principle. The second is the more conservative precautionary principle. And the third, which I will focus on, is the commodification principle. 1

The liberal harm principle

In this traditional liberal approach, activities between consenting adults should not be restricted by law, except to address demonstrable, tangible harms. This principle is often attributed to JS Mill. It places an onus of proof on those who would wish to regulate, and

111 112

113

Rosenbaum, above n 40, 9. Although 2012 presidential candidate Mitt Romney’s odds were inflated by ‘price manipulation by a single trader’, who lost almost US$4 million on Intrade’s market: Rothschild and Sethi, above n 34, 4. On the possibility of betting markets leading to electoral rorts, see Nason, above n 46. Where the law just requires electors to put a mark beside a single candidate.

2014 Betting on Elections 325 _____________________________________________________________________________________

conceives of harm in a limited compass, out of a preference for individual autonomy over other values. Applied to betting, this approach would limit restrictions to those designed to avoid problematic classes of gambling likely to generate tangible harms. An obvious harm is betting by those for whom gambling is so heavy or addictive that it imperils their ability to support themselves or their dependents, psychologically or financially. Legal action on this basis tends to be restricted to measures to educate or empower ‘problem’ gamblers to avoid temptations or to set limits. 114 It is possible some political aficionados could find electoral betting particularly seductive, so that proliferating markets will exacerbate any tendency they have to gamble excessively. But the sporadic nature of elections and the elongated nature of the election betting market mitigate this risk. 115 In contrast, gaming machines are designed to be addictive, whilst racing and sports present a plethora of quick turnaround betting events. In short, under the harm principle, there is little reason for specific regulation of election betting. The harm principle, and its valorising of individual autonomy, however, can be problematic where public goods and activities are involved. Voting and public elections are quintessentially public goods and activities. Selling one’s vote neither harms nor victimises anyone in particular, but it is rightly forbidden because it misconceives the purpose of the franchise and the political duty it represents. The harm principle cannot, therefore, explain the nineteenth century rationale for prohibiting election bets as masking electoral bribery. (A rationale that focused on wagers by candidates or their agents, rather than the modern phenomenon of commercial bookmaking.) 2

The precautionary principle

In this approach, a more cautious attitude to risk is adopted, depending on the gravity of the risk. (The approach is often associated with environmental regulation and systemic risks.) Rules prohibit or deter betting by those assumed to lack capacity, especially minors. Limiting adolescent gambling can be viewed as an application of the precautionary principle, rather than as illiberal paternalism. Despite these restrictions, young people are as likely to experiment with gambling (if lotteries are included) as they are with tobacco or alcohol. 116 Most experience no obvious immediate harm beyond the money lost. But the precaution behind attempts to repress youth gambling is based on population-wide concerns about the relationship between exposure to gambling and propensity for risktaking and mental development. There is no reason to suspect political betting markets hold great attractions to under-18s. Repeated exposure to marketing and gambling odds may desensitise some adolescents to gambling; but political betting odds are of less concern than sporting odds. 117 114 115 116 117

The courts have shown little interest in applying equitable doctrine to save problem gamblers from their contractual obligations: Kavakas v Crown Melbourne Ltd [2013] HCA 25. By ‘elongated’ I mean a book will be open for many months before the electoral event in question. See, eg, Paul Delfabbro et al, ‘Further Evidence Concerning the Prevalence of Adolescent Gambling and Problem Gambling in Australia: A Study of the ACT’ (2005) 5 International Gambling Studies 209. It might be thought that some small exposure is useful, lest gambling markets and odds be constructed as ‘forbidden fruit’ to some adolescents.

326 Federal Law Review Volume 42 ______________________________________________________________________________________

There is however one sub-category of risk where the precautionary principle applies to election betting. That is betting involving political participants like leaders and candidates. Betting by or on behalf of politicians poses conflicts of interest and duty. Rhode and Strumpf cite a report of a senior Australian Minister wagering on the 1949 election; 118 similar reports emerged during the 2010 election. 119 An extreme example would be a political leader betting on whether she would lead her party to the next election. Decisions involving public affairs and politics ought be made for reasons internal to those activities, and not tainted by pecuniary gains offered by betting markets. 120 Such bets also taint the market itself. There is an analogy here with match-fixing in sport. It may be a weak one, as long as political markets are confined to significant events like seat outcomes or leadership contests, where participants have limited control but high intrinsic motivation to succeed. But the analogy is real: books were not suspended on the 2010 election result until long after polling day, even though the question of who would form minority government lay with a handful of cross-bench MPs. 121 A related concern is the corrosion of public trust if there is a perception that representatives are betting on the outcomes of representative democracy. There is a direct analogy here with rules against elite athletes betting on their codes. On balance, a ban on wagering by political participants would be wise under the precautionary principle. It is worth noting that such a ban would not eliminate insider betting, which supposedly adds predictive value to political betting markets. 122 Parties are porous, and information about party polling or caucus sentiment on leadership is available to political insiders like journalists and party administrators. 3

Commodification and political ideals

A final rationale for regulation involves the concept of commodification. The concept is associated today with post-Marxian thought. Yet it shares a conservative, natural law lineage. Its essence is a suspicion of extending free-market liberalism beyond trade in productive goods and services. The suspicion rests on fears that markets have a reductive tendency to reduce human interactions to calculations based primarily on economics or materialism. 123 This concept is problematic to many liberals. Duxbury, whilst not denying that markets may degrade, argues that such concerns do not justify limiting trade in an object or activity merely because some people — even a majority — feel the trade degrades those 118 119 120

121 122 123

Rhode and Strumpf, above n 26 at n 60. Brendan Nicholson, ‘Pollie Punting Crook — PM’, The Australian (Sydney), 2 August 2010, 8. This is not to say candidates cannot back themselves innocently. For an example, see Jon Ronson, ‘Beyond a Joke’, The Guardian (UK, online), 15 January 2000 (a joke candidate backing himself to attract over 250 votes in the hope of recovering the cost of his electoral deposit). Nick Tabakoff, ‘Election Wagers Suspended as Labor Firms’, The Daily Telegraph (Sydney), 7 October 2010, 4. Insider information, as noted earlier, may enhance the predictive value of betting markets; so too may saliency (the presumption that ‘investing’ in a bet encourages punters to research the bet rather than it just relying on hunches). Neil Duxbury, ‘Do Markets Degrade?’ (1996) 59 Modern Law Review 331, 335–6 (discussing the concept of incommensurability).

2014 Betting on Elections 327 _____________________________________________________________________________________

objects or activities. 124 For example, many see pornography and prostitution as commodifying sexuality and exacerbating the sexual objectification of women. The concern is not that any particular example of pornography or prostitution causes harm, but that at some point its widespread availability corrodes deeper ideals. Yet true liberals would baulk at censoring erotica or banning prostitution. An instinct against commodification was apparent in recent debates about the excessive promotion of sports betting during live broadcasts. This led, in 2013, to the television industry code restricting odds during live broadcasts, 125 and governmental measures to ban them altogether. 126 Sport embraces numerous intrinsic values, amongst them tribalism, physical competition and playfulness or chance. Betting per se is not inimical to those values. Indeed wagering on one’s favourite team may be a way of intensifying support, and wagering against one’s team may be a way of hedging disappointment if it loses. Wagers also possess an element of competition (betting against the odds) and of chance, so small-scale wagering is a playful activity. Just note how the US Supreme Court justices run a private election betting pool for their amusement. 127 However when gambling becomes unduly prevalent, it risks overwhelming the practice upon which it is parasitic. Instead of enjoying a match for its intrinsic entertainment, or assessing teams on traditional criteria such as strategy, skill, sportsmanship or the aesthetics of their play, sports betting odds are becoming so predominant that they pollute the native enjoyment of sport. To adapt Wilde, betting may reflect a mindset that knows ‘the price of everything but the value of nothing’. 128 I have an analogous concern with the election betting industry. Politics encapsulates its own complex set of values, notably political participation, equality, liberty and deliberation. Some of these values explain why we would not object to two politicallyminded people having a friendly electoral wager between themselves, or to partisans placing bets to hedge or intensify their election night emotions. 129 With interest in electoral politics waning (as measured by rates of enrolment, turnout and valid voting), we might even hope that election betting offers a way into thinking about politics for some otherwise disengaged citizens. To an optimist then, election bookmaking might yet become an institutionalised part of the ritual of elections, adding a dimension of fun and 124

125 126 127 128 129

Ibid 332. Duxbury’s focus is on trade in unusual goods (eg organs) or activities (eg surrogacy). Such goods or activities are traded in themselves; a betting market is a trade in an invented risk. Liberals would consider markets that are parasitic on other activities to be even less a concern than markets in organs. Commentators cannot now make ‘promotional reference’ to odds during live sporting events, although ‘incidental’ mention of odds and their distinct advertising remain permissible: FreeTV Australia, Commercial Television Industry Code of Practice (2010, as amended 2013), s 8. South Australia is moving to ban any broadcast of odds during a sporting event and to include anti-gambling warnings in sports-betting advertisements or sponsorships. Forrest Maltzman et al, ‘Supreme Court Justices Really Do Follow the Election Returns’ (2004) 37 PS: Political Science and Politics 839. Oscar Wilde, Lady Windermere’s Fan (1892) (defining a cynic as someone ‘who knows the price of everything and the value of nothing’). In a psychological sense, those who wager can have insurable interests, contrary to the formal legal principle that only disinterestedness in the underlying event distinguishes betting from insurance. Compare Goldberg, above n 32, 445–6.

328 Federal Law Review Volume 42 ______________________________________________________________________________________

colour to electoral politics, 130 with few of the addictive risks of betting on high-rotation events such as racing and sports. However if betting markets grow so sizeable that discussion of betting on politics becomes a common adjunct of political discussion — as it has overseas, in times past and present 131 — then there is a commodification risk akin to the problem being experienced in sports. And the loss may be even greater. Sports are just sports: electoral politics is not just a game. 132 Betting on elections was observed, 170 years ago, to configure elections like a horse race or cockfight. 133 Elections already suffer from a proliferation of horse race narratives, drawing on the ubiquity of opinion polling. 134 The deluge of opinion polling provides an easy media commentary that obscures the underlying issues of policy or character that should dominate electoral campaigns. The growth and normalisation of political betting markets risks exacerbating this reduction of politics to the status of a mere game: an infinite regress of reflections on ‘who is electoral favourite?’ rather than ‘who should we favour?’. In a similar vein, Beers observes that during the depression-era general election of 1931 in Great Britain, MPs condemned the electoral majorities market as inappropriate, given the economic and political crisis then enveloping the country. 135 One obvious response is — who cares? Or rather, whose care matters? Cultural sensitivities differ over time, and between people. Life insurance, for instance, faced strong resistance prior to its spread in the nineteenth century, from those who thought it was a profane monetisation of the sanctity of life. 136 Yet if life is sacred, surely it can

130 131 132

133

134

135 136

Compare the portrait of the ‘adventurer’ gambler whose ‘motive is the slipping of perceptual boundaries and moving into the intensity of another reality’: Felicia Campbell, ‘The Positive Case for Gambling: One Person’s View’ in William N Thompson (ed), above n 5, 164, 165, 167. See, eg, Michael Gallagher, ‘The Election as Horse Race: Betting and the Election’ in Michael Gallagher and Michael Marsh (eds), How Ireland Voted 2007 (Palgrave Macmillan, 2008) 148, and Goldberg, above n 32, 439–40. Compare an argument that elections are more ‘sacred’ than the trivial events on which betting has historically focused upon: Matt Beiber, ‘To Bet or Not to Bet? Why Gambling on Elections is Wrong’, The Huffington Post (online), 6 April 2012 . Brewin, above n 24 at 82 (quoting an 1840s newspaper). See also the handbill from the famous Oldham election of 1847, ‘Oldham Races 1847: Great Boro’ Sweepstakes: Value 1600 Suffrages’, giving the ‘Latest Betting and Condition of the Horses’ (ie candidates): reproduced in James Vernon, Politics and the People: A Study in English Political Culture c. 1815–1867 (Cambridge University Press, 1993) 135. Oddly, Leigh and Wolfers argue that we should substitute one form of horse race narrative for another, calling for the media to give greater prominence to reporting of election betting odds: above n 40, 336–7. Similarly unedifying is Goldberg’s argument that more punditry based on election betting could somehow ‘improve campaigns’: above n 32, 438–46. Beers, above n 36, 297. Viviana A Zeliger, ‘Human Values and the Market: the Case of Life Insurance and Death in 19th Century America’ (1978) 84 American Journal of Sociology 591. Prohibitions were common in Europe from the 15th century. Insurance of course involves a kind of betting, on an ‘insurable interest’; true gambling involves an event in which neither party to the bet has any other stake.

2014 Betting on Elections 329 _____________________________________________________________________________________

withstand a little financial hedging. 137 A papal election can attract many bets whilst remaining a moment of solemnity for Cardinals as papal electors and many Catholics alike. However the risk of commodification lies not in the mere fact of any particular insurance or betting contract, but in the size of and attention to the market overwhelming underlying values. This is particularly the case when we are concerned with quintessentially public things such as elections as opposed to innately private things such as bodies. 138 Democracies around the world regulate electoral expenditures, not just out of fear that reliance on large donations corrodes political equality and decision-making, but because an undue focus on political fund-raising obscures the point of campaigning as a communicative dialogue and not merely an advertising free-for-all. A second response to concerns about commodification would be — why fret about election betting? Why not regulate opinion polling? Some countries do regulate polling, by restricting its transmission during the final weeks of campaigning. 139 Aside from free speech concerns about such restrictive laws, a dose of opinion polling may offer two democratic benefits. It can have the deliberative benefit of apprising us of what our fellow citizens are thinking. 140 And it prompts politicians to adjust their positions to take account of popular sentiment. Political betting markets lack either saving grace. Nor, unlike polling (especially exit polling), can betting markets act as a check against electoral fraud. Rather, in their initial framing by bookmakers and to the extent they are dominated by ‘outsiders’ rather than ‘insiders’, election betting markets are derivative upon polling results and the perceptions of favouritism those results generate. Betting on elections then, can be seen as harmless at best, when it occurs on a limited scale and without political participants being involved. But at worst, especially as the industry grows, it represents a potential commodification of politics. This ambivalent status helps explain why most State governments are reluctant to ‘approve’ election betting, despite the interest licensed bookmakers or TABs in those States have in sharing in and expanding that market. Yet State governments also seem uninterested in enforcing (in NSW’s case) or re-enacting (elsewhere) bans on election betting. The reason seems to be that they would risk being seen as draconian, particularly as the laws would be hard to enforce cross-jurisdictionally.

137 138 139

140

Wagers on lives, births and marriages were typical in mid-18th century betting books in London, according to Geoffrey Clark, Betting on Lives: The Culture of Life Insurance in England 1695–1775 (Manchester University Press, 1999) (cited in Swain, above n 50, n 3). I cannot sell my vote, for obvious reasons, even though I might value the money I would receive more than the free exercise of the right. See Article 19, Comparative Study of Laws and Regulations Restricting the Publication of Election Opinion Polls (2003) . The UN Human Rights committee upheld a ban on publishing polling during Korean campaigns to protect a ‘limited period of reflection, during which [electors] are insulated from considerations extraneous to the issues under contest in the elections’: JungCheol Kim v Republic of South Korea, Un Doc CCPR/C/84/D/968/2001 (23/8/2005). Provided it is done scientifically. Some jurisdictions insist on publishing information like sample sizes and margins of error to improve that deliberative benefit: Elections Act 2000 (Canada) s 326 and Publication of Election Opinion Polls Act 2012 (Kenya).

330 Federal Law Review Volume 42 ______________________________________________________________________________________

VI CONCLUSION Historically, both the common law and Australian statutes took a dim view of election betting. Unlike most censorious regulation of betting, this did not reflect class distinctions, such as those between the turf club (acceptable) and the gaming house (forbidden).141 On the contrary, election betting has been as much a pastime of educated or well-off elites as anyone else. Behind those old criminal laws lay a belief that betting introduced undesirable and venal considerations into the electoral process and the voting equation. That is, electors could be influenced to vote a certain way because of a wager. This reasoning applied regardless of whether the wager was between two politically interested persons betting on the candidate that they already favoured (as in the English case of Allen v Hearn) or whether candidates wagered against themselves to induce electors to support them (as in the Tasmanian case of Mugliston v Dillon). The English case was decided prior to the Reform Acts, when the franchise was much less widespread. Whilst this made no difference to the court’s reasoning about the immorality of pecuniary influences on voting, it makes a practical difference. In most modern, mass electorates, a rational elector would feel little incentive to vote one way simply because they had wagered that way (although briberous bets persist in places like Taiwan). 142 More tellingly, rational candidates no longer try to buy votes directly, since it is uneconomic to do so in a mass electorate. So it makes little sense to try to repress election betting out of fear of pecuniary interests corrupting political obligations — except where a politician is involved in the bet. From the viewpoint of democratic participation, there are reasons to be sanguine about small-scale, friendly wagering over elections. Down’s voting paradox holds that in a mass electorate, a ‘rational’ voter has no real incentive to vote. Only other considerations — such as an habituated duty to vote, or the psychological value of self-expression — explain turnout. 143 Whilst the paradox is less obvious in Australia, given the legal compulsion to turnout, it is still relevant as electoral participation is on the wane. Election bets may be a way of enhancing the psychological value of voting. They might even encourage those with a punting mentality to take more interest in electoral politics.

141 142

143

See John O’Hara, A Mug’s Game: a History of Gaming and Betting in Australia (UNSW Press, 1988). O’Hara dates this to the inception of white colonialism, but argues that an egalitarian approach has tended to prevail over-time. Shih Hsiu-chuan, ‘Chiu Yi Alleges Vote-buying under Guise of Gambling’, Taipei Times (online), 10 January 2012 . See also Anon, ‘Legislator Accuses Tycoon of Betting on Poll’, The China Post (online), 10 January 2012 (alleging that approximately A$1m in lost bets would sway 10 000 votes and corrupt a constituency race). Taiwanese election betting otherwise focuses on presidential contests, and was recently repressed for reasons beyond electoral bribery: Tseng Ying-yu and Sofia Wu, ‘Crackdown on Election Gambling to Intensify: Police Official’, Central News Agency (Hong Kong, online), 8 December 2011 . See Geoffrey Brennan and Loren Lomasky, Democracy and Decision: The Pure Theory of Electoral Preference (Cambridge University Press, 1993).

2014 Betting on Elections 331 _____________________________________________________________________________________

The regulatory page, as we have seen, is not blank. In practical terms the genie is out of the bottle, given the spread of the online bookmaking industry. The Commonwealth could, in theory, seek to return the genie to the bottle, given it has power over territories, corporations, telecommunications and national elections. Failing that, State jurisdictions are left with limited regulatory measures. Most have chosen not to licence election betting in their jurisdiction, whilst NSW retains a criminal prohibition on betting on its elections. My position, argued here, is that wagers on elections should be allowed, but not commercially. There is no need for criminal prohibitions, except where politicians bet on events over which they have any say. Friendly wagers are harmless; indeed they can add colour and emotional intensity to people’s experience of elections. They should be lawful, but their enforcement kept out of the courts. Bookmakers, as far as possible, should not be licensed to run books on electoral or political events. Liberals may argue for a commercial freedom to run books on elections. Yet there is no natural ‘right’ to bet on a public artefact such as an election. It is one thing to lay a wager; it is another thing to defend corporate bookmaking and promotion of betting on elections as part of a larger gambling industry. To echo Erskine’s Reflections on Gaming, there is an distinction that should be drawn between ‘gaming as a trade, not as a pleasure; as a mode of living, not of passing life’. 144 The contemporary industry risks a commodification that is antithetical to the public value of elections as moments of deliberative focus and as rituals of democratic participation. 145 Betting markets posit elections as events somewhere between sporting encounters and novelties; not entirely random events, but still essentially games of chance. If unchecked, the growth in election betting markets reinforces a problem inherent in the excessive focus on opinion polling, namely the cynical, reductionist characterisation of elections as mere horse races. Public law, according to Pildes’ version of the expressive theory of law, classically enacts certain values in the face of reductionist claims about rights. 146 Elections are at the heart of public law, and we should be wary of obscuring their meaning by becoming occasions obsessed with financial speculation.

144 145 146

Thomas (later Lord Chancellor) Erskine, Reflections on Gaming, Annuities and Usurious Contracts (Davies, Bew & Walter, 3rd ed, 1777) 4. Election betting embodies a different kind of ritual — the ritual of the risk-taker, checking fluctuating odds before launching into a punt. Elizabeth Anderson and Richard H Pildes, ‘Expressive Theories of Law: a General Restatement’ (2000) 148 University of Pennsylvania Law Review 1503. Contrast Matthew D Adler, ‘Expressive Theories of Law: a Skeptical Overview’ (2000) 148 University of Pennsylvania Law Review 1363. The relevance of the expressive dimension of public law is not only inescapable, but potentially stronger in Australia than the US, where it becomes side-tracked into arguments about constitutional power and rights as trumps.