Finding Furtive Flutters

Peter Van Onselen, definitely one of the rising stars of Australian political science, has a piece in the Bulletin this week on betting markets, which includes this tidbit.

The Bulletin understands that political insiders, armed with internal party research, are dabbling in the betting markets, particularly in seat-by-seat contests. Due to privacy laws, betting agencies cannot disclose the names of individuals who place bets. It is believed, however, that they include sitting MPs, betting both on and against themselves. Profiteering by political insiders has a stockmarket insider trading feel about it, except that it is not illegal.

Over lunch with an election bookmaker recently, I was told about three other groups that frequently bet in these markets: political staffers, union officials (betting on both parties, apparently), and opinion pollsters themselves.

This entry was posted in Australian Politics. Bookmark the permalink.

15 Responses to Finding Furtive Flutters

  1. Which of course would help explain why betting markets have a good record.

  2. Bring Back CL's blog says:

    Insider trading?

  3. Andrew Leigh says:

    Homer, you might have to expand on that.

  4. Bring Back CL's blog says:

    who have first privvy to private opinion polls?

  5. wayaway says:

    I would say there is not a betting market in existence that does not have elements of ‘insider trading’.

  6. Andrew Leigh says:

    Homer, are you repeating something in the post, or making a new point? You have me genuinely bemused.

  7. Bring Back CL's blog says:

    yes just repeating however contrary to popular opinion parties rarely do polling of one electorate because of cost.
    They poll a few marginals. They get this information before the public opinion polls are released by judging by some betting plunges there has been either bad polling or such small polls the MOEs make the figures meaningless.

    Betting on ‘events’ has proved fruitless thus far. punters have picked political events as badly as the commentariat

  8. Andrew Leigh says:

    Betting on ‘events’ has proved fruitless thus far. punters have picked political events as badly as the commentariat

    Nice rhetoric. Now let’s look at the evidence. One of the best head-to-head comparisons we have is this one, from US presidential elections:

    Over the 13 candidacies from 1988-2004, the average absolute error of the market-based forecasts was 1.6 percentage points, while the orresponding number for the Gallup Poll was 1.9 percentage points.

  9. derrida derider says:

    Hmm. It just reinforces what I’ve thought.

    But I’ve concluded that I should always bet on the party I hope doesn’t win, so that if my hopes are dashed I at least get some monetary compensation (mind you, it’ll take a bloody big bet to cheer me up if the Rodent does manage to get back in). The same holds true for sporting contests – I had a flutter against the Wallabies in each of their World Cup matches so my misery at their exit was not wholly unalloyed.

  10. Andrew Leigh says:

    DD, I had a few thoughts on this topic back in June.

  11. Bring Back CL's blog says:


    you have not understood my argument.

    The Commentariat predicted some four /five times the Rudd would be undone by events such as Brian Burke etc.

    Each time the betting market screwed towrds the libs and then the polls would come out and money would come back toward the ALP.
    In other words the ‘smart’ money was not very smart at all.

    I actually took a lot of notice of the Senate seat in Tennessee in the mid-terms. I noticed how much the betting markets reacted to the latest polls

  12. Pingback: Andrew Leigh » Blog Archive » Soapbox

  13. Rajat Sood says:

    On a seat-by-seat basis, I think the betting agencies show (or at least showed recently) that the Coalition will be returned. Assuming that the seat-by-seat bets are more efficient than overall bets due to insider trading, does this mean that the Coalition has a better chance than it seems?

  14. Andrew Leigh says:

    Rajat, I don’t think that’s right. Simon Jackman tells me you have to look at his “expected seat count”, rather than his “share of seats >0.5” count. If it weren’t so, there’d be an enduring arbitrage opportunity.

  15. Bring Back CL's blog says:

    by the way the reason for the discrepancy between the national betting and seat by seat betting is easy to explain.
    Parties do not do individual seat polling. They poll a group of marginal seats.

    I would say Bennelong is different this time though.

    So the betting in individual seats is more gut feeling more than anything else

Comments are closed.