Will the Polls Be Wrong Again On the Democratic Primary?

opt_IMG_6845Boy, polling has had a rough go of it the last few years.  First, they vastly underestimated Obama’s 2012 margin of victory (thanks Gallup).  Then they were significantly Democratic leaning in 2014.  Combined with international polling misses in Israel, the UK and domestically with the Kentucky guberntorial race and you can see why polling misses get noticed more and more.

But, none of that can compare to the polling miss last week in Michigan.  According to the RCP average of polls not a single poll had Clinton leading the state by single digits.  Yet, when the final votes were tallied she lost narrowly.

The reasons offered for the miss but they seem to focus around the idea that Clinton damaged herself with trade.  Sander’s focus on the issue probably hurt the wife of Mr. NAFTA.  But that still does not explain why the last three polls up to election day had Clinton ahead by 27, 37 and 24.

I doubt we will get a good answer but exit polls showed Sanders crushed Clinton among whites and lost blacks 60-40.  This suggests Midwestern blacks, more educated, unionized and middle class, are more receptive to the Senator than Southern blacks (just look at Mississippi).

Next week most of the focus will be on the GOP battles for Florida and Ohio but the Democratic races across the country will be interesting.  Illinois, Missouri, North Carolina, Ohio and Florida vote.  Polling has been sparse in each state but Illinois and Florida stand out.

Polls show Clinton with massive leads in both and yet both places have fertile populations for Sanders support.  Nobody expects a Michigan upset in either but that is exactly the problem.  The polls could be well off the mark.  In Florida, Sanders has a wealthy, liberal elite to poach and a young Cuban and Puerto Rican base to draw from.  In Illinois, it is hard to see Cook County blacks acting differently than say Wayne County blacks.

The same dynamic holds true for North Carolina and Ohio.  Blacks make up a sizeable chunk of the Democratic base in both states but so do young, college educated whites as well as rural, unionized whites (more so Ohio).  Polling in Michigan significantly underestimated Sanders strength among the latter.

Now, a caveat is in order here.  There are few good pollsters in Michigan but national outfits also thought the race was a foregone conclusion.  Well, few good outfits have polled Illinois or Ohio.  Heck, Missouri and North Carolina for that matter.  Clinton will probably add to her delegate total and is the odds on favorite to win all 5 states.  But, after Michigan, I don’t have much faith in the polls.



What to Watch For on Super Tuesday GOP Style

635920304493623629-AP-GOP-2016-Debate.1Both Hillary Clinton and Donald Trump are in commanding positions heading into Super Tuesday.  Buoyed by victories in recent weeks both front-runners expect to dominate the Southern heavy voting.

But straight up victories will not tell the story of who won the night.  The reason why is simple: delegates.  Specifically, the way the state and national parties allocate them.  Due to RNC and DNC rules, no state is allowed to allocate all its delegates to the winner until after March 8th.  In an effort to get around this rule though many Super Tuesday states have minimum threshold support a candidate must maintain to receive delegates.

These rules have the potential to make the ultimate winner of Super Tuesday, well, less of a winner.  Tomorrow, voting states will allocate their delegates proportionally with a threshold of either 15 percent (Arkansas, Oklahoma) or 20 percent (Texas, Alabama, Georgia, Tennessee, Vermont) of the vote. In other words, the state divides up its delegates proportionally between the candidates who surpass the threshold.

For some candidates these rules are incredibly beneficial (Rubio and Cruz).  Yet, for others, Kasich and Carson, the rules are disastrous because neither is polling strongly outside of one or two states.  For Trump, the rules stink because it means his wins will not allocate him all the delegates of the states.

However, this is not a given.  Cruz and Rubio have the potential to see their lanes of support split with Kasich and Carson in the mix.  This means that in some states they could fall under the needed thresholds for delegates.  The primary beneficiary of this would be Trump but it also is true that Cruz or Rubio could benefit (assuming one falls under the threshold).

Winning/losing and pure percentages are not the only things that matter.  So will margins.  For example, if Cruz is racking up big margins in TX it likely means he is winning congressional districts and tallying up more delegates.  By the same token, smaller Trump victories tomorrow mean he is likely not winning all the districts he could have.

Put another way, according to RNC rules, each congressional district gets three delegates to the national convention. And in many states (including Texas, Georgia, Alabama, Tennessee, among those voting Tuesday) those delegates are allocated based on the results in each district. If a candidate wins a state by a large margin, he’s more likely to rack up delegates in more of that state’s congressional districts. We saw this in South Carolina – Trump beat Rubio 32.5 percent-22.5 percent, but that was enough to carry every congressional district and net him all 50 of the delegates.

Most notably, in the two biggest delegate rich prizes tomorrow, TX and GA, a majority of delegates are allocated by CD.  In TX it’s 108 (36*3) while in Georgia it is 42 (14*3) out of 155 and 76 total delegates.

For some candidates some states have more meaning than others. Cruz desperately needs to win TX because be badly trails Trump in many states and Rubio is edging him out in others.  In addition, Cruz could probably never recover losing his home state.

Trump, due to his commanding lead in delegate totals and polls to date, can easily stand to lose Texas and still rack up the most delegates for the day.  Even a massive Cruz win in TX would likely not blunt Trump’s margins elsewhere.

For Rubio, the strategy is simple to survive until Florida.  If he can finished second in many states (VA, AR, OK, GA, MN and AL) are probably his best bets he can plausibly come out of Super Tuesday still in the game and the candidate of the establishment.  If you are Carson or Kasich, it’s a wing and a prayer night.