Rick Perry and the Latino vote, part 1

Say what you want about Rick Perry, he’s got a much firmer grasp of the changing demographics of Texas and their political implications than many of his partymates do. As such, he plans to compete vigorously for the Latino vote in Texas.

Perry campaign manager Rob Johnson said the campaign will try and improve upon the one-third of the Latino vote that the governor has won in past elections.

“We can do better,” he said.

And for the future of the party, it must do better, Johnson said, citing that Hispanics will make up 50 percent of the state’s population in 10 years. A party that only wins one-third of that vote will have an uphill election battle, he said.

I’ll leave it to others to judge the efficacy and likelihood of success for Perry’s strategy. What I want to do is check Rob Johnson’s math. I don’t have access to any exit polling data from 2002, so I’m going to do my best to take a rough guess at Perry’s support level among Latinos from 2002 by looking at State Rep district data. What I’ve done is pulled out all of the Governor’s race returns from the SRDs in which the percentage of Spanish surname voter registrations (SSRVs) is at least 50. Here’s what that data looks like:

HD  Representative   Perry   Sanchez Perry% Sanchez%  SSVR
==========================================================
031        Guillen   1,965    18,154    9.8     90.2  91.0
033           Luna  12,466    16,167   43.5     56.5  53.1
034        Capello  13,861    14,512   48.9     51.1  52.4
035        Canales  15,794    17,186   47.9     52.1  52.2
036         Flores   4,857    13,168   25.3     74.7  79.7
037       Oliveira   4,833    10,360   31.8     68.2  81.7
038          Solis   7,465    11,614   38.8     61.2  74.5
039           Wise   5,288    12,417   29.9     70.1  78.1
040           Pena   2,829    11,678   19.5     80.5  86.8
041      Gutierrez   9,137    10,516   46.5     53.5  62.6
042        Raymond   3,399    27,357   11.1     89.9  85.5
043        Herrero   9,615    14,975   39.1     60.9  68.2
074        Gallego  13,998    16,342   46.1     53.9  55.7
075    Quintanilla   5,541    11,940   31.7     68.3  77.7
076         Chavez   3,659    16,769   17.9     82.1  64.6
077         Moreno   4,640    13,191   26.0     74.0  69.9
078       Haggerty  14,662    12,041   54.9     45.1  51.6
079        Pickett   6,815    10,759   38.8     61.2  80.0
080          Garza  11,179    15,628   41.7     58.3  68.9
104         Alonzo   3,650    10,342   26.1     73.9  51.3
116 Mrtinez-Fischr   8,858    12,798   40.9     59.1  57.5
117         Mercer   9,748    10,621   47.9     52.1  56.7
118         Uresti   9,907    11,917   45.4     54.6  56.7
119         Puente   8,781    13,061   40.2     59.8  57.2
123     Villarreal   9,927    13,325   42.7     57.3  58.2
124       Menendez   9,052    12,551   41.9     58.1  57.0
125         Castro  11,861    13,973   45.9     54.1  58.0
143         Moreno   3,890     9,294   29.5     70.5  58.3
145        Noriega   4,500     9,921   31.2     68.8  60.4

Total              232,177   392,577   37.2     62.8

There are many caveats to keep in mind. Not everybody who has a Spanish surname is actually Latino, and not everyone who is Latino has a Spanish surname. The percentage of Spanish surname voter registration may or may not have any relation to the percentage of people with Spanish surnames who voted. And even if you assume that the share of Latino voters is more or less constant from county to county and district to district, there will still be fluctuations. Basically, we’re using a yardstick to measure molecules.

Having said all that, I don’t see anything in these numbers to contradict what Johnson said. I do want to note, however, that the more heavily Latino a district was, the worse Perry tended to perform. To put it another way, I added up the vote totals in all the districts where the SSRVs were in the 50’s, in the 60’s, in the 70’s, and 80 or above. Here’s what that looks like:

SSVR Pct             Perry   Sanchez Perry% Sanchez%
====================================================
50-60              146,455   184,130   44.3     55.7
60-70               42,730    81,000   34.5     65.5
70-80               23,151    49,139   32.0     68.0
80+                 19,841    78,308   20.2     79.8

Again, this metric is too crude to make any strong conclusions, but the trend is clear enough. If I were Nate Silver I’d draw you a graph and throw some correlation coefficients at you, but let’s just pretend I did that for now.

I did not do the same analysis for the other statewide candidates, because there are only so many hours in the day. My eyeballing of the data suggests that in most places Perry did better than some of his ballotmates, like David Dewhurst, and worse than others, like Carole Keeton Rylander. In other words, with the exception of HD42, which is in Tony Sanchez’s home county (Webb), Perry’s relative position was about where it was overall. I did not see anything that suggested to me that he did better than you might expect. Maybe I’ll tackle that another day.

The conclusions I will draw are that Perry is certainly capable of getting a third or better of the vote in heavily Latino areas, and that if his efforts aren’t matched by something at least as strong, he will do well enough to make a Democratic victory all but unattainable. You’ll see more evidence of that in Part 2.

Related Posts:

• Abbott and the Latino vote
• On polls and turnout
• What it takes to win statewide