In three previous articles I looked at how the polls for the Democratic nominee for President compared to the bookmakers’ odds, which in turn reflect what the betting public believes will happen. This is an update to see how the landscape has changed after the Iowa caucuses and New Hampshire primary.

A snapshot taken of the polls and odds on five dates presents a good summary of what has happened: (i) August 15, when Elizabeth Warren’s odds first passed Joe Biden’s, (ii) October 13, when Warren’s odds peaked, (iii) November 20, around my last report, and (iv) January 24, right before the Iowa caucuses, and (v) February 18, with results from the first round of polls after the New Hampshire primary. All data are taken from averages compiled by

First, the national poll numbers, in %:


  Aug. 15 Oct. 13 Nov. 20 Jan. 24 Feb. 17
Biden 30.5 29.4 30.7 28.9 16.5
Warren 17.3 23.2 18.0 14.6 12.8
Sanders 16.0 16.6 16.7 22.7 27.3
Harris 8.0 5.0 4.3
Buttigieg 5.2 4.4 8.0 7.4 10.8
Booker 2.0 1.6 1.3
Yang 1.5 2.6 2.3 4.0
Bloomberg 3.0 7.6 15.8
Gabbard 1.0 1.3 2.0 1.4 1.7
Klobuchar 1.4 1.5 1.7 3.6 6.8
Steyer 1.4 1.0 2.1 1.8


Next, the odds, in % (anyone who has never been above about 5% not shown):


  Aug. 15 Oct. 13 Nov. 20 Jan. 24 Feb. 18
Warren 29.6 53.4 26.8 11.3 2.0
Biden 27.9 19.9 25.5 35.6 11.4
Harris 16.6 4.5 2.2
Sanders 14.1 6.3 15.7 33.6 41.8
Buttigieg 8.6 7.7 20.5 7.2 11.4
Yang 5.0 5.9 5.6 4.5
Clinton 2.5 8.7 7.0 4.5 5.7
Bloomberg 4.9 12.5 32.1
Klobuchar < 1 < 1 2.5 1.5 5.0

(Note: the odds’ percentages add up to more than 100% since these are inferred from betting odds, which have a vig for the bookies cooked in.)

Biden of course was the big loser in the polling after Iowa and New Hampshire, as his showing indicated he is not really what the (Dem) people want, at least in those states. Bloomberg was the big gainer, despite not even being on the ballot in New Hampshire. Buttigieg gained a little (surprisingly little since he essentially tied with Sanders in the first two contests), and Warren lost a little, and continued her downhill slide. Klobuchar’s numbers went up after better-than-expected performances in Iowa and New Hampshire.

In the odds, Biden was the biggest loser, as the bettors seem to mostly think that he is done. However, that might be a bit of an over-reaction (and markets can over-react, right?) – more on this below. Warren’s assumed chances also have taken another major hit. Bloomberg was the biggest gainer, followed by Sanders and Buttigieg (the latter still not back to his November peak, however). Clinton still has a few who believe she has a chance (cue ominous music).

On the horizon we have the Nevada caucuses (Feb. 22), the South Carolina primary (Feb. 29), and then Super Tuesday (Mar. 3), which includes big-voter states California and Texas. The most recent average poll results in these states are shown below, along with the date of the most recent poll for each (state polls are done much less often than nationally).


  Nevada South Carolina California Texas
  Feb. 16 Feb. 13 Jan. 27 Feb. 9
Biden 16.0 26.5 21.0 28.0
Sanders 30.0 20.0 25.8 22.7
Warren 14.5 9.0 19.8 15.0
Steyer 10.5 16.0 2.7
Buttigieg 12.5 7.5 7.3 5.7
Klobuchar 9.5 4.5 3.5 3.3
Bloomberg 4.3 11.7


Biden appears to have a lot of strength still, although only two of the polls here are post-New Hampshire and California is pre-Iowa. Nevada and South Carolina have similar numbers of electoral votes compared to Iowa and New Hampshire (6 and 6 versus 6 and 4), so Biden could challenge for second place in the delegate count before Super Tuesday. Sanders also looks to do well through Super Tuesday.

The most striking aspect of this to me is the support for Bloomberg in the odds, without corresponding support in election results or the polls. Apparently a fair number of people are willing to put down money that Sanders is not going to get the nomination, and they seem to think that among the other candidates Bloomberg has the inside track. My read is that even if Sanders was ahead in the delegate count going into the convention, due to the fragmented field he wouldn’t have a majority, and the non-Sanders delegates would coalesce around some other candidate. The “smart money” may believe that candidate would be Bloomberg, perhaps because Warren is too much like Sanders policy-wise, Biden is, well, Biden, and Buttigieg doesn’t have as much experience. The only other choice among front-runners is Klobuchar, but perhaps she also lacks experience and name recognition, although her midwest connection could help in some swing states (the Dems are winning New York no matter who is running).

Another factor in Bloomberg’s popularity with bettors may be his perceived electability. The table below shows how various candidates do versus head-to-head Trump in recent polls. They are ordered by the bias they have in favor of the Democratic candidate, which is pretty consistent across all candidates.


  Bloomberg Biden Sanders Klobuchar Warren Buttigieg
Quinnipiac +9 +7 +8 +6 +4 +4
CNN +9 +9 +7 +3 +5 +4
FOX News +8 +9 +6 +1 +5 +4
LA Times/USC +9 +7 +2 +4 +3
NBC/WSJ +7 +8 +4 +3 +3 +4
NPR/Marist +4 +6 +3 +2 +1 +2
ABC/WashP +3 +4 +2 +1 0 -3
Emerson 0 +2 0 -2
IBD/TIPP +1 +1 -2 -4 -3
Average +5.9 +5.9 +4.1 +2.5 +2.0 +1.4


It’s interesting that the FOX polls are similar to CNN’s, whereas Investor’s Business Daily’s are the most favorable for Trump. Also, Biden appears to do as well as Bloomberg.

If Sanders leads the delegate count going into the convention (my understanding is that superdelegates don’t vote on the first ballot), and doesn’t ultimately get the Dem nomination, it could get interesting. If that were to pass, here’s a sneak preview of the popcorn futures market on the morning of the Democratic national convention (from an Oct. 23, 2012, article by Anthony Watts on

As I said in the last post, there is much more detail at, if you want the complete picture (of polls and odds, not popcorn futures).