Wednesday, May 18, 2016

Prediction markets and understanding politics

A post in a closed group on Facebook suggested that we look to prediction markets for information on politics, and try to use the information they offer in the manner that business journalists use financial markets. 
     The trouble is that business journalists know as much about business as sportswriters know about sports and political journalists know about politics. Fortunately, students of politics have it easier than students of sports or business because politicians have to explain themselves to get what they want. If you want to understand Trump, the first step is to look to his own words, the second step to look to words of the politicians who are his allies, and the third to look to the words of the politicians who are his rivals and enemies.
     As for prediction markets, the efficient markets hypothesis says they integrate publicly available knowledge. That does not mean that they are accurate, just that anything more accurate is going to cost you.

     The odds on a Republican winning the Presidential election (getting a majority of the two-party vote) are about 1 in 3 at the Iowa Electronic Market .  If I know better, if I know that Trump is a virtual lock because he is the more centrist of the two candidates and because the country is generally perceived to be on the wrong track, it may be that that is because I have been studying politics for 30 years.
     Should you trust me? I have not, so far, put any money down on Trump at Iowa. But the people who bought Trump in the Republican convention market at IEM are doing well so far, as today's graph shows.

Are the prediction market odds a good guide to Trump's actual chances?  Were they a good guide to his chances the day after the Iowa caucuses?  All we can say with any degree of assurance is that they are a better guide than any other you can get for free.  
    Trusting me isn't free.

Friday, May 13, 2016

No, Trump is not a menace to the Constitution

Josh Blackman, a con law professor who supported Rand Paul and Ted Cruz, claims in the National Review that "Trump has already promised that he will knowingly break the law and violate the Constitution," because his views on constitutional matters are not exactly the same as Blackman's.  But one can be a perfectly good constitutionalist and hold the Times v. Sullivan is bad law (it is certainly inconsistent with the Founders' understanding of the relation between state libel laws and the First Amendment.) Likewise with Kelo, which simply means that judges should defer to the judgment of elected officials on what constitutes public use, And, finally on Japanese internment, to quote Judge Posner, "Korematsu was rightly decided."
      As for the rest of Blackman's complaints, no doubt President Trump will exercise his constitutional powers to try to move things his way, whether that is by executive action, speechifying, proposing legislation, or having the Trump Administration's lawyers make the appropriate arguments in the courts. Trump's promises on constitutional matters are best understood as statements of direction and principle rather than marching orders. That is how candidates in the US always talk, as if they are going to be able to by themselves what they have to move the whole apparatus of government and public opinion to accomplish.