Four or five in the frame and a chaotic process makes it far too hard to call

The Iowa caucus is one of those things that no-one in their right mind would invent if it didn’t already exist. As a democratic process, it flouts goodness knows how many rules of good practice, from high barriers to participation to non-secret voting to a lack of consistent process. However, it’s possible to get too precious about process. The caucus does exist and has survived, whatever its flaws, because it tends to work.

What’s the purpose of the primary process if not to select a candidate widely acceptable to the party and as well-placed as possible to win the general election in November? For all that most members of the public can participate on one side or the other, these are, after all, internal party elections. In that sense, the caucuses are not all that unlike the CLP meetings currently nominating candidates for Labour’s leadership and deputy leadership elections.

This matters because the caucus process makes them notoriously difficult to poll. Not only is each election cycle different in terms of candidate spread, public interest, whether both or only one party has a natural vacancy and so on, but it’s part of the purpose of the meeting for activists to try to change voters’ minds. Further, the caucus meetings use a quasi-Supplementary Vote system whereby supporters of ‘non-viable’ candidates can recast their votes in a second round. The threshold for viability varies according to the size of the district, from 15% for most to 50% for those with only one delegate. Working out who will attend and who they’ll end up supporting is a far more complex task than, say, polling the November election.

That redistribution point matters because there are still eleven candidates in the field (following the withdrawal of John Delaney yesterday), four of whom – Biden, Buttigieg, Sanders and Warren – are generally polling mid-teens or better, with Amy Klobuchar also heading that way. However it’s probable that in most districts one or two of those five will miss the initial threshold, as well as the six also-rans, meaning maybe 30% or so of votes could be recast.

(We should note that missing the threshold first time round isn’t automatically an eliminator – activists can still try for their candidate to reach it second time round with enough net transfers)

What all this should hopefully show is how much uncertainty there is in the process. Not only is the field relatively closely spread in the polling averages – for reference, the RCP average at the time of writing was Sanders 23.8 / Biden 20.2 / Buttigieg 15.8 / Warren 14.6 / Klobuchar 9.6 – but those polls themselves are all over the place. The spread of the last four polls reported on RCP has:

Sanders 19-30% (1st or 2nd places)
Biden 15-25% (1st to 4th places)
Buttigieg 10-18% (3rd to 5th places)
Warren 11-19% (2nd to 4th places)
Klobuchar 6-16% (3rd to 5th places)

The most watched-for poll – that of the Des Moines Register – should be published tomorrow evening. While that does have a good record from previous elections, it still the same difficulties of methodology to contend with as other polls and I’d be very wary of putting too much trust in it. For reference, the 2016 polls gave Hillary Clinton a 3% lead (she ended up winning by fractions), and on the Republican side the DMR poll gave Trump a 5-point lead whereas in the actual caucus, Ted Cruz won by 3%.

All of which says to me that this is a race well capable of springing a surprise. Sanders is rightly favourite but the best odds available (8/11, Ladbrokes) are far too short. I don’t think anyone is an odds-on shot.

Sanders’ narrow (first-preference) lead doesn’t sit well for me with what I suspect is not great transfer-friendliness unless Warren seriously under-performs. I’d have thought that Biden (5/2), Buttigieg (10/1) and Klobuchar (33/1) all offer more value.

David Herdson

p.s. Next week will be a busy one in US politics. After the Iowa caucuses on Monday, Trump will deliver his State of the Union address on Tuesday and then the Senate will vote on his impeachment on Wednesday (and almost certainly vote to acquit unless he says something extremely stupid in his State of the Union speech, which may be more overtly political than usual). For our purposes here – looking at the caucus outcome – the important thing is that the impeachment vote will happen afterwards: there should be no major change in the political situation between now and when the caucuses are held.