I want to repost something I wrote after Nevada. Statistically, Hillary "won" following Iowa
Considering the "path" for Bernie has all but shut on his underwhelming performance tonight, it is time for me to break out an oldie but goodie (God, I do enjoy getting this kind of stuff right).
-----
So some of you who frequent GDP are probably aware that I write primers on polling, statistics, etc when there is incorrect or misunderstood information out there. I haven't done a real one of these in some time because a) we're actually voting now and b) they're not being met favorably or as non-partisan in GDP as the campaign heats up.
Considering this entire post is going to be how, numerically, Hillary is pretty much a foregone conclusion...well, I'm not going to even pretend there is a reason to post this in GDP other than for being an asshole. This seems like the much more appropriate place to post.
My standard disclaimer, in real life I am a statistical analyst...blah, blah, blah.
There are two kinds of analyses that are regularly talked about on boards like these; those attempting to show the numbers and those that attempt to explain what the numbers mean. This second type, if done without directly citing from a singular numerical analysis, is called a narrative analysis. A narrative analysis is one where readers are invited to look up the numbers for themselves while explaining the story that those numbers show. This will be a narrative analysis because it is dependent on a wide assortment of collected data and not just one or two polls.
One of the things many people have issues with is being made into a number. I get it; we don't want to see all that makes us us boiled down to a few metrics. It is dismissive of the human condition, and it is really infuriating when those metrics turn out to be accurate.
In 2008, statistical analysts knew Clinton was going to lose after SC. There was still a lot of votes to be cast, chances for small upsets etc, but people who look at numbers could tell by the demographics breakouts that Clinton was going to come up short. I was a Clinton supporter in 2008, and I still voted for her in the NY Primary, but I knew it was a lost cause. Obama's strengths as evident in the states he won (and lost) early on pointed to the end outcome with an amazing degree of confidence.
It took three states for analysts to reach that conclusion in 2008. This year it took one.
Everyone knows the logic behind this. Bernie hasn't broken through with minorities, with older voters, with registered Democrats, with women, etc, but Iowa was seen as the best indicator if Bernie's base of support would be enough to overcome those significant disadvantages. And, to be honest, Bernie significantly outperformed expectations in Iowa...but he underperformed versus what was needed to win it all. And not just by a little.
See, despite what some people here would say, caucuses heavily favor what is perceived as Bernie's base. Caucusers tend to have a much higher percentage of self identified activists than primaries. And Iowa is overwhelmingly white and allows independents to caucus. So the caucus make up included the three strongest legs of Bernie's base (activists, white, Democratic leaning independents). This was the litmus test to see if the Bernie coalition could overpower the amazingly broad base that Clinton has built.
Now, even a win by Bernie in Iowa wouldn't have been enough for analysts. From sites like Cook's and 538 as well as others, it was estimated that Bernie needed to win 60 to 70% of the delegates in Iowa to overcome the Hillary train via demographics and the more intangible "momentum" shift. We all know that this didn't happen.
People who do work like me (or better than me) saw this as the beginning of the end for Bernie. Now some couched these beliefs by saying he could make up ground elsewhere, anything could still happen, etc. For me, considering this is a message board and that I do have a dog in the fight, I'll say nope. He is done, barring Hillary having to drop out.
And weirdly enough the aftermath of the giddy Bernie supporters celebrating NH confirmed the numbers were going to work against him. As Super Tuesday polls began rolling in, it was clear that the demographics in those states rejected the momentum notion of the Bernie campaign. His victory in NH did not resonate in the more diverse states where Hillary had built up quite the base. Again, for analysts this was predictable based on the supporter make up of each candidate, but numbers are very cold and unfeeling about such things. The problem for people accepting this is that passion will almost always outweigh analysis.
We all know what happened in NV. Clinton won the moderates, the older more reliable voters, the women, the African Americans and, yes, the Latinos (Best estimate I have seen so far is that she won the Latino vote by a 16 point spread). Hillary's base is so strong that a challenger can't syphon off enough to break all the way through. The Bernie coalition was never big enough to do it, not even in Iowa. The numbers rarely lie, the demographic support is simply not there to lead an insurgency campaign to the candidacy itself.
Bernie is still going to win some states, we all know that, and Hillary is still going to have to work her tail off for awhile longer, and all of us, regardless of who we support, need to get out there and exercise our right to vote. But math says it is time to start bringing this home. I'm not saying Bernie should drop out of anything silly like that. But Hillary is about to rack up a lot of delegates, and she is more than likely going to start pivoting some of her rhetoric to the Republicans after Super Tuesday. And why shouldn't she? The odds are well and truly in her favor.
= new reply since forum marked as read
