Just because Hillary won some some majority Hispanic precincts doesn't mean anything. Hispanics aren't some monolith, and enough Hispanics elsewhere in the state could have voted Bernie for him to win them statewide.

I'm sure the 28% who are Hispanic are spread out throughout the state. No need to look at a handful of precincts. Hispanics aren't a monolith where those 22 precincts (which include non Hispanics as well) are going to vote the same as Hispanics statewide.

1. The entrance/exit polls were flawed.

2. The caucus system apportionment of delegates in Clark County precincts with a high number of Latino voters does not accurately model the actual Latino popular vote for the entire state of Nevada.

3. A combination of 1 & 2.

I'd say that possibility 2 is far more likely than possibility 1, but factor 3 cannot be ruled out (with factor 2 being the greatest factor in explaining the poll vs. highly Latino Clark County precinct delegate discrepancy).

Here's why. Entrance and exit polling simply asked self-identified Latinos who they planned on voting for and recorded the results. While sampling models are only as accurate as how representative they are of the entire population they are attempting to model, there is nothing anyone has presented about the methodology used that would explain why it was clearly unrepresentative.

On the other hand, assuming the allocation of delegates of highly Latino Clark County precincts is more representative of the individual voter preferences of the Latino population of the entire state of Nevada requires a myriad of tenuous assumptions:

1) the caucus apportionment of county delegates actually accurately reflects the popular vote of each precinct.

This is an extremely tenuous assumption, especially for precincts that award a small number of delegates. The apportionment for each precinct was preselected and had nothing to do with the actual turnout at each precinct. In addition, if say 6 delegates were to be awarded for a certain precinct a 3 to 3 apportionment for each candidate could represent anything from 58.3% popular vote for Sanders to a 58.3% popular vote for Clinton.

2) the caucus apportionment of county delegates actually in "heavily Hispanic" neighborhoods accurately reflects the popular vote of Latinos in these neighborhoods.

This is an even more tenuous assumption, for both the reasons outlined above, and more importantly because we have no information on the voting preferences nor the relative turnout of any non-Latino voters residing in these precincts.

3) Latinos who live in "heavily Hispanic" Clark County neighborhoods are representative of all Latinos who live in Nevada.

This is yet another extremely improbable assumption because it assumes that the already tenuously assumed voting preferences of the subset of urban Latinos who live in highly heavily Hispanic neighborhoods in just Clark County is actually representative of the voting preferences of all Latinos in Nevada regardless of the cultural, racial, generational, linguistic, and economic demographics of their actual neighborhoods as well as their population density, agricultural and industrial profiles.

because the numbers for that subcohort were so small the results were meaningless. Of course one candidates supporters decided to claim that this meant the overall poll was bogus or didnt contain any folks from that age group.

This analysis by Nate Cohn makes sense It depends how many hispanic voters were interviewed and from where.

He said exit polls are no good but then he included the exit poll that showed Sanders losing among black voters.

They're wrong except when they're right?

He said that the numbers for small minorities would be so small out of a 1024 person poll that the results could easily be off for those subgroups depending on a number of factors. Since Hillary won heavily hispanic areas, it makes sense that the suggestion that she lost the hispanic vote is suspect. She also won the areas with the most African American voters bolstering the polls suggestion that she won the African American vote.

she won way more delegates in those precincts than that poll suggests. Now, to be fair, I am not sure the reasoning is totally sound but I do agree the Hispanic numbers are likely to have been off. It is hard to see her doing that well in delegates from Hispanic precincts while losing the vote. In the case of the AA numbers the fact she won them going away could well lead to that lopsided of a delegate split.

they put an end to the minority unity meme.

Bernie supporters said he did win the Hispanic vote so doesn't that mean it's true??? I mean, they're never wrong - right?

What is the use of getting powerful as minorities if we become the same sell outs as those who were in power. This is a sad loss for dignity and for the weaker members of my community. Pretty disgusted, actually.

When Worlds Collide

http://www.dsausa.org/cornel_west_and_dolores_huerta

Sid

Thanks bigtree.

On a side note, it's entirely possible for one to white and Hispanic or black and Hispanic. Would love to see the Hispanic vote broken down more but that may be asking too much of the pollsters.

Excellent for Hillary!

then find out you lost it before they leave.

I read this...