Did Susana Martinez win at least a plurality of the Hispanic vote?

[redcommander]

I know Sandoval didn't, but wasn't sure about her.

[JohnnyLongtorso]

I'm going to say no. Of the counties with a high Hispanic population (Guadalupe, Mora, Rio Arriba, and San Miguel, all in the 70-80% Hispanic range), Martinez got 57%, 46%, 41%, and 38%, respectively. Doesn't look like she won a majority of Hispanics, since she presumably did win white voters.

[ag]

Well, there is some negative correlation between the Hispanic share of population and her vote share, so, probably, she didn't, especially considering that not all Hispanics are citizens. But, still, she was quite impressive (her weakest areas are mostly urban, Native or liberal, not Hispanic).

So, just going by county (Hispanic % by 2000 census):

Mora 81.64% Hispanic, 45.6% Martinez
San Miguel 77.96% Hispanic, 38.4% Martinez
Rio Arriba 72.89% Hispanic, 40.6% Martinez
Gudalupe 81.22% Hispanic, 57.2% Martinez
Doña Ana 63.35% Hispanic, 51.7% Martinez
Taos 57.94% Hispanic, 29.4% Martinez
Luna 57.7% Hispanic, 58.4% Martinez
Hidalgo 56% Hispanic, 58.3% Martinez
Valencia 54.98% Hispanic, 59.9% Martinez
Santa Fe 49.04% Hispanic, 32.9% Martinez
Grant 48.79% Hispanic, 48.7% Martinez
Socorro 48.73% Hispanic , 52.8% Martinez
Colfax 47.49% Hispanic, 60.2% Martinez
Harding 44.94% Hispanic 62.3% Martinez
Chaves 43.83% Hispanic 70.3% Martinez
Bernalillo 41.96% Hispanic, 51% Martinez
Lea 39.65% Hispanic, 74.1% Martinez
Eddy 38.76% Hispanic, 69.1% Martinez
Quay 37.98% Hispanic, 65.1% Martinez
Torrance 37.15% Hispanic, 67.2% Martinez
De Baca 35.27% Hispanic, 68.9% Martinez
Union 35.1% Hispanic, 72.8% Martinez
Cibola 33.42% Hispanic, 48.7% Martinez (note: this one is 40.32% Native American)
Roosevelt 33.29% Hispanic, 71.4% Martinez
Otero 32.16% Hispanic 69.6% Martinez
Curry 30.38% Hispanic, 72.4% Martinez
Sandoval 29.4% Hispanic, 56.5% Martinez
Sierra 26.28% Hispanic, 65.7% Martinez
Lincoln 25.63% Hispanic, 73.3% Martinez
Catron  19.16% Hipanic, 77.5% Martinez
San Juan 14.99% Hispanic, 69% Martinez
McKinley 12.4% Hispanic, 34.85 Martinez (note: this one is 74.72% Native American)
Los Alamos 11.75% Hispanic, 54.3% Martinez

[cinyc]

By my math, Martinez averaged 49% in the 9 majority-Hispanic counties and 62% in the rest (in both cases, not weighing for population).  Seems like a fairly close call - but probably not.  Without doing a precinct level breakdown of Bernalillo County, I guess we won't know for sure.

We can certainly conclude she lost among American Indians and government employees in Santa Fe, though.

BTW - It's kind of bizarre that there were no exit polls of New Mexicans or Rhode Islanders this cycle.  I guess without a Senate race, the media didn't want to pay for one.

[phk]

Likely not.

-.6791208*(%Martinez) + .8232409
-.6791208*.5342 + .8232409 = 0.460454569*100 => 46.05%




. reg  PercentHispanic PercentMartinez

      Source |       SS       df       MS              Number of obs =      33
-------------+------------------------------           F(  1,    31) =    8.45
       Model |  .251451794     1  .251451794           Prob > F      =  0.0067
    Residual |  .922579606    31  .029760632           R-squared     =  0.2142
-------------+------------------------------           Adj R-squared =  0.1888
       Total |   1.1740314    32  .036688481           Root MSE      =  .17251

------------------------------------------------------------------------------
PercentHis~c |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
PercentMar~z |  -.6791208   .2336365    -2.91   0.007    -1.155626    -.202616
       _cons |   .8232409   .1397648     5.89   0.000     .5381888    1.108293
------------------------------------------------------------------------------

[Nichlemn]

This poll suggests Martinez got 38% of the Hispanic vote.

I am suspicious of that site's consistently high Democratic numbers, though. Firstly, when your polling is different from everyone else's, it's either because you're right and everyone else is wrong, or you're wrong. All else equal, the single outlier is probably wrong.

Secondly, if other exit polls have consistently underestimated Hispanic Democratic support, one of two things must happen. (a) Exit polls must consistently be biased towards Republicans. The 2004 Presidential exit polls are a glaring counterexample. (b) Other exit polls also underestimate white Republican support. You need a story for this as well, and while you could make one up (say, underpolling of rural areas), it seems rather implausible for there to be errors balancing each other.

Finally, it conflicts with analysis of localised voting patterns like phknrocket1k's above. True, the difference isn't that large and it might be caused by say, whites voting more Republican in more Hispanic areas (contrast effect?) But there are other results like those in Texas which seem hard to reconcile. If Rick Perry is only winning 21% of the Hispanic vote, how did Republicans win Hispanic majority TX-23 and TX-27? Even given the large non-voting Hispanic population, they're still very Hispanic, especially the latter (72%). Maybe there's a lot of ticket splitting, or Hispanics/whites in these districts are considerably more Republican than they are in the state at-large. But the simplest explanation is that Hispanic voters are in fact not all that Democratic.

[Dgov]

This poll suggests Martinez got 38% of the Hispanic vote.

I am suspicious of that site's consistently high Democratic numbers, though. Firstly, when your polling is different from everyone else's, it's either because you're right and everyone else is wrong, or you're wrong. All else equal, the single outlier is probably wrong.

Secondly, if other exit polls have consistently underestimated Hispanic Democratic support, one of two things must happen. (a) Exit polls must consistently be biased towards Republicans. The 2004 Presidential exit polls are a glaring counterexample. (b) Other exit polls also underestimate white Republican support. You need a story for this as well, and while you could make one up (say, underpolling of rural areas), it seems rather implausible for there to be errors balancing each other.

Finally, it conflicts with analysis of localised voting patterns like phknrocket1k's above. True, the difference isn't that large and it might be caused by say, whites voting more Republican in more Hispanic areas (contrast effect?) But there are other results like those in Texas which seem hard to reconcile. If Rick Perry is only winning 21% of the Hispanic vote, how did Republicans win Hispanic majority TX-23 and TX-27? Even given the large non-voting Hispanic population, they're still very Hispanic, especially the latter (72%). Maybe there's a lot of ticket splitting, or Hispanics/whites in these districts are considerably more Republican than they are in the state at-large. But the simplest explanation is that Hispanic voters are in fact not all that Democratic.

It also could matter where they take the poll.  Urban, Border, and Poorer Hispanics tend to vote far more Democrat than the Suburban and middle-class ones.  However, the Poor/Urban Hispanics tend to be more concentrated in one area (Urban slums), which is why most of the 90% Hispanics precincts vote like 75-80% Democrat, even as Hispanics statewide are only like 60-65%.