June 14 Poll Report

            Good morning everybody.  Yesterday’s poll—Biden, Booker, Buttigieg and Warren—was the strongest group of candidates I have offered in at least three weeks, controlling an estimated 44% of the vote, and that number clearly growing.   The poll went essentially as would have been predicted by previous polls, except that Elizabeth Warren gobbled up half of Cory Booker’s rapidly diminishing voting stock, and also picked up trace amounts from Biden and Buttigieg.  Here is the chart summary of that poll:

            It is clear, it has been clear for several weeks now, that the race is consolidating around a limited number of leaders.    The standard deviation of scores for the candidates, which hit a low of 316 on May 27, has been growing almost every day since then, and is now up to 402.   The late-June Democratic debates could either reverse that trend, if some of the struggling candidates do well, or (more probably) emphasize that trend, by effectively pushing some of the candidates off the stage.     A total of 386 points changed hands in yesterday’s poll, a large number as was expected since it was a power poll.   Because it was a big poll and did shake up the standings, I am going to take a few minutes to summarize the initial and secondary effects of the poll:

            Elizabeth Warren picked up another 89 points yesterday, in essence another 1%.  She now stands at 1733, easily the strongest position that anyone has had since I started these polls in early April.   I wouldn’t make too much of this.   We don’t know to what extent Warren’s support in my polls does or does not reflect her support in the wider electorate, and, even if it does, 17% is not 50%.  And even if she is in reality well ahead of her rivals, there is a long time left for her to stumble.  

            Joe Biden, despite not doing especially well in yesterday’s poll, picks up 58 points, mostly due to the removal of the April 24^th poll from the data.  Biden got 54% in the April 24 poll against Trump, Castro and Swalwell, but indications are that if we repeated that polling group now, Biden would do better than that.  

            Pete Buttigieg, who performed as expected in yesterday’s poll, picks up 31 points as a secondary effect of the forward steps of Biden and Warren, who he was being polled against yesterday. 

            Kamala Harris advances 9 points from yesterday, as a secondary effect; I’ll explain better in a minute.

            The biggest thing that happened yesterday was actually not Warren over-achieving by 6 points in the poll; it was Cory Booker under-achieving by 5 points.  Given where they are, Booker losing 5 points—half of his support—is a heck of a lot bigger event than Elizabeth Warren picking up another 6. 

            Explaining how the polls work with an analogy to college basketball, suppose that Kentucky has a rating of 800 and Kansas a rating of 900, and suppose that Kansas and Kentucky play in late January in the Big 12/SEC challenge, which they always do.   Suppose that Kentucky wins.   In that scenario, Kentucky’s victory helps not only Kentucky, but, to a lesser extent, helps everyone in the SEC (Kentucky’s conference) and hurts everyone in the Big 12 (Kansas’ conference.)  If Kentucky has lost earlier to Tennessee but Kentucky then beats Kansas, that makes Tennessee look even stronger.  

            I poll the candidates in "heats" pretty often, polling Cory Booker against other candidates who have similar strength ratings—Bernie Sanders, Amy Klobuchar, Beto O’Rourke, Kirsten Gillibrand, etc.   When Booker has a terrible day, that lowers the "values", the "position estimates", for everyone that Booker has been polled against.  It makes his entire conference look bad.   So, as a result of yesterday’s poll, not only does Cory Booker lose 37 points, but everybody in his group loses a few.  Beto O’Rourke loses 8, Andrew Yang, Bernie Sanders, John Hickenlooper, Jay Inslee, Kirsten Gillibrand and Andrew Yang lose 7 each, Tulsi Gabbard loses 6 and Jeff Flake loses 5.   Amy Klobuchar loses 14, and Stacey Abrams loses 13.  

            But not all of that is a result of Booker’s bad day; some of it is a secondary effect of Donald Trump steadily falling in my polls.   Republicans are not voting in my polls in representative numbers.   I regret this; I wish that they would, but their candidates just lose and thus look bad, so I understand logically why they would not wish to participate.  

            The poll removed yesterday was the April 24^th poll, in which Donald Trump got 26%.   The removal of that poll knocks Trump backward just as far as yesterday’s poll knocks Booker backward, 39 points—and Trump is at a relatively similar strength number to Booker.   He’s in the same conference.   So it is like Kansas loses to Kentucky and, on the same day, Oklahoma loses to LSU; the whole conference’s reputation takes a hit.   That’s why all of those candidates are down about 2% each.  

            It’s not a big deal; Klobuchar and Sanders and Gillibrand will probably recover the few points they have lost in the next few days, O’Rourke and Hickenlooper and Abrams will probably continue to fall a few points.  

            Julian Castro also lost 15 points since yesterday as a result of the removal of the April 24^th poll.  In that poll Joe Biden got 54%, Julian Castro 15%, a ratio of 3.6 to 1.   Their current numbers are 1192 and 202, a ratio of almost 6 to 1.   So that poll suggested that Castro had some reasonable level of support vs. Joe Biden.  Pulling that poll out of the data costs Castro points.

            These are the updated standings: