During an election campaign, when we want to know the likely outcomes, we normally look to the five main voter opinion polls that are published by the mainstream media. But as Shaun McGirr and Rob Salmond argue in their chapter in the new post-election book Key to Victory: The New Zealand General Election of 2008 (edited by Stephen Levine and Nigel S. Roberts) there are some problems with these established opinion polls, and in the 2008 election there were ‘two emerging challengers: “polls-of-polls” and election prediction markets’. Their chapter asks ‘which sources of political information New Zealanders should trust in election campaigns’?, finding that ‘some of the options are substantially superior to others’. [Read more below]
Political scientists Shaun McGirr and Rob Salmond sum up their chapter with the following questions: ‘which of the many “predictive” mechanisms used in New Zealand in 2008 did the best job. Did the prediction markets live up to their promise? Did aggregating five smaller polls into on large poll produce a net gain in accuracy? Did some polls continue to display their historical biases?’ (p.262). This is a lot of ground to cover, but the chapter’s examination of political prediction markets is the most interesting element here.
Political prediction markets
The most significant political prediction market is iPredict, which is apparently ‘run by a wholly owned subsidiary of the Victoria University of Wellington and the Institute for the Study of Competition and Regulation’ in which ‘traders ventured real money to buy a variety of futures (contracts tied to particular outcomes)’ (p.258). As explained by iPredict, 'In their simplest form, stocks on iPredict pay $1 if an event comes true, and nothing if it doesn't. If stocks are selling at 60c, it means the market believes there is a 60% chance of that prediction coming true. If a buyer thought the odds of the prediction coming true were higher, they might buy a lot of stock and this would drive up the price'.
Some significant amounts of money were invested in 2008:
- $64,500 was traded over the likely nature of ‘the Maori Party’s post-election relationship with National’
- $25,800 was traded over the Wellington Central battle between Grant Robertson and Stephen Franks
- $132,100 was traded over whether ‘there will be a National prime minister after the 2008 election’
- $413,000 in total was invested in election-related predictions
McGirr and Salmond go into very useful detail to explain the concept of prediction markets. They refer to James Suroweicki’s landmark 2004 book The Wisdom of Crowds as part of the growing recognition that the collective wisdom of a well-informed crowd can be very accurate in making predictions.
In prediction markets, members of the public buy and sell contracts tied to the outcome of some future event. If there are many diverse buyers and sellers in markets where people trade on whether an outcome will occur – for example, that “there will be a National prime minister after the 2008 election” – if Suroweicki is right, then the price of these contracts should converge on the true probability of the event actually occurring. When contracts are about a quantitative result – “the Green Party’s percentage of the party vote”, for example – the contract price would converge on (or close to) the true result. If these markets are truly efficient, then the actual price of a contract will also be the price that maximizes expected profit. This price will be more accurate than many other prediction methods because the market mechanism encourages research by market participants and action based upon the information gathered. As with polls-of-polls, prediction markets aggregate the information available from a wide variety of sources to produce a more accurate result than any individual’s prediction (p.258).
What is especially powerful about predictive markets is their potential use of what might be seen as ‘insider trading’:
Markets encourage the discovery and also disclosure (by buying or selling contracts at a higher or lower price) or information, especially private information, by rewarding participants with greater returns for acting on that information. However, traders who end up with low quality private information will eventually be punished by the market mechanism when their prediction is revealed to be false (p.260).
Of course such ‘insider trading’ could be seen as a flaw in the predictive market model, as some traders might be motivated by attempts to distort the prediction market thereby furthering their own political interests. But as McGirr and Salmond point out, predictive markets have a strong incentive to self-correct due the financial rewards involved:
One interesting element of prediction markets is that not ALL players need to be rational – Hollywood studios and election candidates alike can elevate the price of their stock by purchasing it at higher and higher prices. But their joy is likely to be short-lived as rational players capitalize on the opportunity to buy or sell a mispriced contract for a quick profit (p.261).
Nonetheless, there is seen to be a bias in political predictive markets that actually use real money in the transactions. McGirr and Salmond say that ‘real-money markets tend to be populated by wealthier people’ and thus ‘it should not be a surprise that a real-money market will be biased rightwards’ in their predictions (pp.267-268). But this is, of course, good news for those leftwing traders who can seek to benefit from this distortion by making more profitable trading of shares.
So, how accurate was iPredict in 2008? McGirr and Salmond conclude that although iPredict overestimated the eventual support for both Labour and National, it was more accurate any individual polling company.
Political opinion polling
There’s some discussion of the various biases of different polling companies, due to their different sampling methodologies. McGirr and Salmond argue that in theory TVNZ’s Colmar Brunton poll and the NZ Herald’s DigiPoll probably overestimate right-leaning support, and that Roy Morgan poll has a slight leftward bias (p.257). In reality in 2008, McGirr and Salmond found this to be the case – with Colmar Brunton and DigiPoll exaggerating public support for National, and Roy Morgan exaggerating support for Labour (p.264).
So which polling companies were most accurate and inaccurate? McGirr and Salmond say that TV3’s TNS poll was the best (as it was in 2005 as well), and Fairfax’s Neilson pool was the poorest.
In addition to the five opinion polls, some observers attempted to average out the idiosyncratic errors of the individual polls by aggregating them into a “poll-of-polls” using different methods. The New Zealand Press Association simply took the average of the estimates of the six most recent polls, while The New Zealand Herald took the average of the last four polls. Two blog-based polls-of-polls – one run by David Farrar of New Zealand’s permier political blog Kiwiblog, and one hosted at a smaller blog [run by author Rob Salmond] called 08wire – weighted more recent polls with larger sample sizes more heavily (p.257).
McGirr and Salmond say that ‘Poll-of-polls consistently performed well during the 2008 campaign, outperforming most of the opinion polls and the prediction markets’ (p.270). They therefore advocate that both the media and public should pay much more attention to this highly accurate source of political information.
UPDATE: Matt Burgess of iPredict has commented on the chapter on the iPredict blog.
Table of contents
Preface - Stephen Levine and Nigel S. Roberts
Overview of the Election
2008: Key to victory - Stephen Levine and Nigel S. Roberts
2008: The last baby-boomer election - Colin James
2008: Leadership during transition - Jon Johansson
Political Party Perspectives
National - Steven Joyce
Labour - Grant Robertson
The Greens - Catherine Delahunty
ACT - John Boscawen
The Maori Party - Rahui Katene
The Progressives - John Pagani
United Future - Rob Eaddy
New Zealand First - Damian Edwards
New Zealand’s party system: a multi-party mirage? - Jennifer Curtin and Raymond Miller
2008: Images of political leadership in the campaign - Claire Robinson
2008: Media coverage of the election - Babak Bahador
2008: The international media and the election - Aljoscha Kertesz
2008: The campaign in cyberspace - Nicola Kean
2008: The YouTube campaign - Rob Salmond
2008: Voting behaviour and the keys to victory - Stephen Levine and Nigel S. Roberts
2008: The impact of the Electoral Finance Act - Bryce Edwards
2008: Opinion polls and prediction markets in New Zealand - Shaun McGirr and Rob Salmond
2008: National’s winning strategy - Therese Arseneau
Key to Victory: The New Zealand General Election of 2008
Levine, Stephen (ed)
Roberts, Nigel (ed)
Victory is the
story of the New Zealand general election of 2008, in which the experienced and
long-serving prime minister, Helen Clark, was ousted by a political newcomer –
National’s John Key.
Veteran academic commentators Colin James, Jon Johansson, and Therese Arseneau offer perspectives on what New Zealanders were voting for when endorsing John Key and National, and what they were voting against. Several MPs elected for the first time in 2008 provide first-hand accounts of their parties’ campaigns, including Labour’s Grant Robertson; the Greens’ Catherine Delahunty; the Maori Party’s Rahui Katene; ACT’s John Boscawen; and the director of National’s winning campaign, Steven Joyce, appointed to Cabinet following National’s victory. New Zealand First’s doomed campaign is described by its campaign director, Damian Edwards, while party strategists John Pagani and Rob Eaddy provide accounts of the Progressive and United Future campaigns.
Key to Victory also investigates the important issues of the 2008 election, such as the impact of the Electoral Finance Act, and the likely future of New Zealand’s remaining small parties.
During the 2008 campaign political parties started getting to grips with websites, blogs, Facebook and YouTube, and ‘prediction markets’ competed with traditional polls in forecasting the election results. The book describes these developments and provides insights into the use of the media by John Key and Helen Clark in their rival campaigns for leadership. International reaction to the New Zealand campaign and the country’s vote for change is also highlighted.
Key to Victory includes a special DVD with excerpts from key campaign events including the televised leaders’ debates, the leaders’ opening night campaign addresses, parties’ TV ads and campaign billboards.