State of election markets: 394 Days

The market was pretty calm this week with Jeb Bush continuing to hold off Marco Rubio as the most likely candidate. His polling numbers look good, but his funding is still lagging; with several month left until the first voting, a poll of money is a little more important than the poll of the people. In the lower tiers, Carly Fiorina continued to cool off a bit with a small decrease in likelihood of victory, as Donald Trump and Ben Carson refused to budge (too much). But, the most interesting market news did not occur in any market, but in polling. Gallup declared that it would run no horse race polling for the 2016 primary. Polling has officially surrendered to markets!

Sources: Betfair, Hypermind, PredictIt, http://twopointoh.predictwise.com/politics/2016RepNomination

Bush and Rubio continue to provide about 60% of the total likelihood of victory with Bush holding the slightest of edges. On one hand Rubio is looking stronger than Bush in the polls with 7, 10, 8, 13, 11 in the last five polls on Huffington Post’s Pollster to Bush’s 14, 7, 7, 10, 7. But, at the same time Rubio raised just $6 million in the third quarter of 2015. This means that the establishment is not quite ready yet to open up their wallets and surge Rubio to victory.

Fiorina began fading this week from 8 or 9 to 6% likelihood of winning the nomination. There was speculation that she could surge up into a Donald Trump vacuum or take non-establishment support from Ben Carson. But, reports of Trump’s demise were premature and Carson has held steady as well (and is fundraising like crazy). This puts Fiorina in a difficult spot somewhere between an insurgent non-establishment candidate and an establishment candidate. She needs to find that niche soon to get into the hunt to win the nomination.

In 1936 George Gallup famously launched the best known survey research firm in the world by accurately predicting the presidential election; in 2015 Frank Newport announced that Gallup will not produce polling for the 2016 primary. Newport cited a desire to focus on polling the electorate on issues, which is great; affecting issues is why we have elections. But that is not the reason they would drop horse race polling; horse race polling is an advertisement for a multi-million dollar survey research company. Elections are the only major polling domain where we have something that resembles a ground truth (with the obvious caveats of changes over time). Gallup correctly realizes that “probability” polling is likely to do even worse in the next few elections and Gallup does not want to destroy its reputation with any pending disasters (i.e., Gallup does not want to pay for its own negative advertisement). It is probably a good public relations move for Gallup, but it bodes very poorly for traditional survey research if Gallup no longer wants to stand up and be tested.

If you want some objective data on the 2016 primary I have a suggestion that is not hiding away from the difficult: prediction markets.

The Democrats will host a debate on Tuesday, October 13. Better late than never?

Sources: Betfair, Hypermind, PredictIt, http://twopointoh.predictwise.com/politics/2016DemNomination