Romney's comeback was Hispanics

[King]

In 2008, a lot of people were talking about lower conservative enthusiasm or more women for Obama yadda yadda yadda.  But to me, the shift from a 2004 Bush 51-48 win to a 2008 46-53 Obama win came mostly in one obvious sweeping change of a demographic: My demographic.

Hispanics moved from 54-45 Kerry to 67-31 Obama.   Hispanic voting has been increasing election to election.  It's hard to say exactly what a 13% swing from Republican to Democrat was in volume.  It's believed about 15 million voted in 2004 and 16 million 2008.  Even if Obama won all 1 million new voters, that's still just 56-44.  McCain had to have lost a lot.

67% of 16 million is around 10.7 (+2.6) million for Obama to 5.3 (-1.5) million for McCain.  So, mathematically, one could estimate that 1.5-2.0 million 2004 Bush Hispanic supporters voted for Obama. 

If Kerry had done as well as Obama with Hispanics in 2004, he would have won the popular vote.  At 252 a combination of Hispanic rich New Mexico, Nevada, Colorado (19) or a win in Florida (27) would have also won him the electoral college.

So, let's talk Obama-Romney with Hispanics.
It was kind of a foregone conclusion that Romney had lost this demographic early.
70-21 were polls at the beginning of the summer with this group.  It actually looked like he would do worse than McCain.  To me, and I said it many many many times this year, it meant Romney no shot at winning this election. 

And, for most of the year, he hasn't had a shot.  Since the debates, however, he has had a shot to win this election.  So, what happened? 

I haven't really thought about Hispanics until today.  I just kept assuming Obama was blowing them out and the reason Romney got back into the race was increased Republican enthusiasm among whites and all those other generic reasons.  Obama's camp seems to think that, too.  There's another thread on the forum right now about how they think 72% white projected electorate will doom Romney.  But I happened to notice something in today's new PPP national poll:

PPP 10/25 Obama +1


37%.  Mitt Romney is closing back in on Bush era 40%+ Hispanic support.  Outlier?

Politico 10/18 Romney +2



Even better for Romney.  Two outliers?  Unlikely.  What's going on?  How long has it been like this? Let's see if we can spot the pre/post-debate trend.  Unfortunately, there's no PPP or Politico data on this I can find before October.  Here's SurveyUSA instead:

SurveyUSA 9/16 Obama +7



That's Romney favorability on the right.  -39 head to head and -31 in opinion.  Really, really bad numbers pre-debates, but then again a lot of people with no opinions formed.

Post-debates...

SurveyUSA 10/22 Romney +3



We're an impressionable people.

[Ty440]

Latino Decisions , the gold standard of Latino polling says otherwise.

[Ljube]

Hispanics care about the economy. The economy isn't doing well. Four more years of Obama won't improve the economy.

Logical conclusion - give Romney a chance.

What effect is this going to have in Nevada?

[rob in cal]

   King, those numbers are fascinating.  I'd love to see a really detailed breakdown of the hispanic vote, by country of origin, how many generations in this country, marital status, religion etc..
   I heard somewhere that Hispanic protestants are much more Republican than Hispanic Catholics.  I wonder if this is because lots of nominal, but not church attending, Hispanics are Catholic, and Hispanic protestants are more active in their churches and take in a more conservative social ideology from that.  Or, is that a sign of greater assimilation into the US society.
    Also, I'd love to see a vote breakdown by residence.  The Hispanic, mostly Mexican descended kids who grew up in my area, suburban Sacramento, might very well lean Democratic, but many of them take on attributes, cultural patterns, aspirations and ideas similar to their white friends and neighbors.
    I'll be interested to see how the Hispanic vote goes on the statewide propositions this year in California.  Its starting to be prime time for the "coalition of the ascendent" to be churning out some electoral victories out here, so we'll see what happens.

[King]

Latino Decisions , the gold standard of Latino polling says otherwise.

Well, 1 being right against 4 is unlikely.  An advantage it does have is that it is a full poll which targets Hispanics specifically, while these are subsampled.  Subsample margin of error shouldn't really account for 20-30% shifts, though.

If Obama is still really in the mid-60s with Hispanics, then that means Obama's support in these polls are understated.  If Obama gets back into the mid 60s in Hispanics who turnout, he'll win.   If Romney can touch 40, it's quite possible he will win.

A movement of Hispanics more than any other group could also explain why Romney has not made huge inroads in Ohio (only 3% Hispanic) during this bump yet Colorado, Florida, Nevada are starting to move closer.  The most recent California poll shows Romney at 41%. Even Virginia and North Carolina are 8-9% Latino states, far more than Ohio.

[Ty440]

Latino Decisions , the gold standard of Latino polling says otherwise.

Well, 1 being right against 4 is unlikely.  An advantage it does have is that it is a full poll which targets Hispanics specifically, while these are subsampled.  Subsample margin of error shouldn't really account for 20-30% shifts, though.

If Obama is still really in the mid-60s with Hispanics, then that means Obama's support in these polls are understated.  If Obama gets back into the mid 60s in Hispanics who turnout, he'll win.   If Romney can touch 40, it's quite possible he will win.

A movement of Hispanics more than any other group could also explain why Romney has not made huge inroads in Ohio (only 3% Hispanic) during this bump yet Colorado, Florida, Nevada are starting to move closer.  The most recent California poll shows Romney at 41%. Even Virginia and North Carolina are 8-9% Latino states, far more than Ohio.

ll agree with you , if Romney hits 40% he will most certainty take Colorado, Florida and come close in Nevada and win the popular vote.

Chuck Todd was right , he said on Morning Joe the other day that Hispanics may just"save" Obama

[5280]

These numbers basically say it's a Bush 2004 running against Obama 2012.

[bgwah]

Interesting analysis.

]The most recent California poll shows Romney at 41%.

I do think it's worth noting that pollster's previous two polls showed Romney at 40% and 39%. Not a big change.

Also, don't forget polls showed Republicans picking up NV and CO in 2010, and that didn't happen.

[old timey villain]

I just don't see any reasonable explanation as to why the GOP is doing better with Hispanics, if this is accurate. Anti immigrant vitriol doesn't win over Latino voters.

This is what Latino Decisions is reporting, for what it's worth

http://www.latinodecisions.com/blog/2012/10/24/ld-vote-predict-indicates-latinos-voters-will-support-obama-over-romney-by-3-to-1-margin/

[Bacon King]

Those are all robopolls, right? I've noticed a huge disparity between the Hispanic subsamples of robopolls and live-caller pollsters this election, with the former consistently showing Hispanics as much more Republican.

This inaccuracy was visible in the primary polls as well- I remember PPP in Florida consistently had Gingrich winning like 45% of Hispanics, who of course voted overwhelmingly for Romney.

[Southern Senator North Carolina Yankee]

Those are all robopolls, right? I've noticed a huge disparity between the Hispanic subsamples of robopolls and live-caller pollsters this election, with the former consistently showing Hispanics as much more Republican.

This inaccuracy was visible in the primary polls as well- I remember PPP in Florida consistently had Gingrich winning like 45% of Hispanics, who of course voted overwhelmingly for Romney.

Which pollster are robopolls besides PPP. They were the only ones that had Gingrich winning Hispanics in the FL primary, if memory serves me. It is likely why they so underestimated Romney's win there.

Another thing to consider is that if Romney were only getting 20% of the Hispanic vote, he couldn't even be close in FL and he clearly is.

[angus]

We're an impressionable people.

All polls citing "hispanics" are suspect.  If you're hispanic, then you know that race means something different to them than what the white anglophone and black anglophone gringos agonize over.  It usually means culture more than anything.  But every ten years, the Census Bureau spends millions of dollars trying to box all of us up.  Generally, whitefolk and blackfolk play their silly game, but the Census Bureau understands that hispanics do not.  They try to offer it in different terms every time, as well--compare the 1990, 2000, and 2010 forms--trying to get an accurate count.  Currently, their best estimate is that something like 18 million US hispanics (out of around 50 million) checked the "other" box.  They simply do not view themselves the way the pollsters would like them to view themselves.  Don't try to pick apart the difference between 31% and 37% and 44% when it comes to "hispanic support for Republicans."  It's a futile effort.

[afleitch]

I posted this a few days ago:

The biggest survey of it's kind, Latino Decisions http://www.latinodecisions.com/blog/2012/10/22/latino-voters-highly-engaged-in-presidential-election-support-for-obama-rises/ has been conducting a large scale tracking poll in both English and Spanish for Hispanic/Latino voters in America. In most national polls Hispanic's are a small subset of voters; at time's no more than 50-100 people polled. However these small sample sizes can impact on the reported poll results.

Latino Decisions as of today found that 71% said they would vote for Obama. 20% said they would vote for Romney. During the aftermath of first debate 'bounce' the figures were 67-23. The Pew Hispanic Center http://www.pewhispanic.org/ in their slightly older poll had Obama at 69% to 21% for Romney. Over at Gallup their September 24th-October 14th 'super sample' that had Obama up 3 in Registered Voters overall had the Hispanic vote at 69-26 for Obama.

EDIT: Their sample from Oct1-21 found the Hispanic vote to be 69-27; a similar gap.

So in these three large comparable surveys Obama leads Romney by 51, 48 and 43 among Hispanics.

If we allocate 68% to Obama and 22% to Romney (the average of all three figures) just for fun then Obama leads by 46 amongst Hispanics. Exit polls in 2008 gave Obama a 36 lead. So Obama appears to be doing better this time around.

What happens therefore if we remove the small samples of Hispanic voters from the national polls and drop down the above figure? This doesn't for example tinker with likely voter screens or implied turnout. It just corrects for small errors in the data amongst Hispanic voters.

Let's have a look at the PPP poll from 18th-20th October which gave us a draw with both candidates at 48%. Obama lead Hispanics by 57-35, making up 9% of the Likely Voters in a sample of 1200 voters. So theoretically, Hispanics made up around 108 responses. Obama got 62 responses and Romney 38. If we adjust this to the figues I suggested above (which remember are based on larger samples and in the case of two reports; directly designed to gauge voting intention amongst a minority group) Obama 'gains' 11 responses and Romney 'looses' 14 responses. Doesn't seem like much. PPP rounded their figures so we have to work with 576-576. If we adjust these then Obama gets 587 and Romney 562.

That gives Obama 48.9% to 46.8%. That's 49-47 for Obama. Such a small change in the subsample but a bigger change in the headline figures.

What about the IDB/TIPP poll that gives Obama a 5.7% lead? Well they don't have figures for what % of the electorate Hispanics represent, but do give a voting breakdown 62-24. Obama's lead in that poll would be higher. What about the Politico poll today which has Romney ahead 49-47? Well they reported 44% vote for Romney amongst Hispanics (which would be the highest % vote a Republican president has ever got with Hispanics). If we re-weight that poll then we end up with an Obama lead of 49-47

Now of course small errors in the other demographic subsets could cancel out these errors. This model is simply designed to highlight potential pitfalls in measuring support in one key group

If the Hispanic vote is as decisive for Obama as Latino Decisions, Pew and even Gallup suggest, then Obama is in a stronger position that it perhaps appears. Given that white voters make up 70+% of voters, then the sample sizes are large enough to reduce potential error. What we don't have of course are wider surveys for African/American and other minority voters.

[SetonHallPirate]

Those are all robopolls, right? I've noticed a huge disparity between the Hispanic subsamples of robopolls and live-caller pollsters this election, with the former consistently showing Hispanics as much more Republican.

This inaccuracy was visible in the primary polls as well- I remember PPP in Florida consistently had Gingrich winning like 45% of Hispanics, who of course voted overwhelmingly for Romney.

Which pollster are robopolls besides PPP. They were the only ones that had Gingrich winning Hispanics in the FL primary, if memory serves me. It is likely why they so underestimated Romney's win there.

Another thing to consider is that if Romney were only getting 20% of the Hispanic vote, he couldn't even be close in FL and he clearly is.

Florida's hispanics are a different breed from hispanics anywhere else in the country, due to the Castro Factor.

[Torie]

How anyone can poll the Hispanic preference when their turnout this year is so up in the air, escapes me. I suspect Hispanic turnout will substantially erode from 2008.

[opebo]

How anyone can poll the Hispanic preference when their turnout this year is so up in the air, escapes me. I suspect Hispanic turnout will substantially erode from 2008.

Why would it do that?

[tmthforu94]

What an interesting analysis - thank you for sharing, King.

With national polls showing Romney getting a substantial bump among Hispanic voters, perhaps there won't be the election day bounce for Obama in Southwestern states that many are projecting. 

[You kip if you want to...]

Well, he is an anchor baby himself, so er...

[All Along The Watchtower]

How anyone can poll the Hispanic preference when their turnout this year is so up in the air, escapes me. I suspect Hispanic turnout will substantially erode from 2008.

Why would it do that?

You must be logged in to read this quote.

http://politicalticker.blogs.cnn.com/2012/10/25/latino-vote-may-top-12-million-in-2012/

[opebo]

I suspect Hispanic turnout will substantially erode from 2008.

Why would it do that?

You must be logged in to read this quote.

http://politicalticker.blogs.cnn.com/2012/10/25/latino-vote-may-top-12-million-in-2012/

Yeah, the voter suppression strategies by the GOP and low registration could dampen the increase of Hispanics as a share of the electorate, but it is hard to imagine it could (given their enormous increase over four years) actually hold them to levels below 2008.

[Sbane]

How anyone can poll the Hispanic preference when their turnout this year is so up in the air, escapes me. I suspect Hispanic turnout will substantially erode from 2008.

Was it even up in 2008? I think this is just a media meme that all non-whites turned out in droves to elect Obama when it's not really true at all. Black turnout was certainly up, but Hispanic or Asian turnout wasn't any higher than what you would expect normally. Of course it will go up since a lot more kids are Hispanics or Asians than 75+ year olds who die and don't vote. And of course those who are naturalized within those 4 years between presidential elections.

[Torie]

We shall see. Hispanics are just so tough to poll. They are a mystery within an enigma wrapped within a riddle. One just has to deal with that, or engage in self torture. Be well.

Oh, I might add, that Dubya's share of the Hispanic vote in 2004 has been placed anywhere from about 33% to 44%, and that is after all the votes have been counted, and then analyzed to death. That is just how tough it is.

[Bacon King]

How anyone can poll the Hispanic preference when their turnout this year is so up in the air, escapes me. I suspect Hispanic turnout will substantially erode from 2008.

http://cis.org/projecting-2012-hispanic-vote-nationally-battleground-states

This analysis suggests Hispanic turnout will increase considerably from 2008, up to 8.9% of the electorate (from 7.4% in 2008). It's worth noting that this same methodology accurately predicted 2010 Hispanic turnout, within a tenth of a percent.

[BaldEagle1991]

How anyone can poll the Hispanic preference when their turnout this year is so up in the air, escapes me. I suspect Hispanic turnout will substantially erode from 2008.

http://cis.org/projecting-2012-hispanic-vote-nationally-battleground-states

This analysis suggests Hispanic turnout will increase considerably from 2008, up to 8.9% of the electorate (from 7.4% in 2008). It's worth noting that this same methodology accurately predicted 2010 Hispanic turnout, within a tenth of a percent.

With this data so far I think Obama will win Nevada and Colorado.