Despite Conservative Bias, Republican Poll Still Shows Disastrous Disadvantage for Republicans Among Latinos

On Friday, the GOP campaign consulting firm Ayres, McHenry, and Associates released a new poll sponsored by the Resurgent Republic—a GOP research and public affairs organization—and the Hispanic Leadership Network, an outreach effort of the right-of-center American Action Network.  Both groups work for the election of additional Republicans to Congress and the defeat of President Obama.

The poll, widely reported in the National Journal and other papers, purports to show significant weakness of President Obama among Latinos in important swing states of Colorado, Florida, and New Mexico.

The central finding, that the President is currently underperforming his 2008 support among Latinos in those states, is consistent with our recent work nationwide.  Nevertheless, in our view, the poll most clearly illustrates two things: 1) that amateurs with no familiarity with the Latino electorate frequently make large and methodologically important mistakes in how they poll; and 2) that the Republican brand is so severely compromised among Latinos—even in Florida—that it may be their undoing in the 2012 election.

First, Ayers/McHenry reports conducting a registered voter poll of Latinos (a sample frame that we use at Latino Decisions), and yet report only 1-2% of all interviews are conducted in Spanish.  Since 40% of Latino adult citizens are foreign born and speak Spanish as a first language, this number should be greeted with extreme skepticism.  Though Ayers/McHenry reports that respondents had a choice of language, that very low representation of Spanish-preferring respondents demonstrates significant methodological problems in how they afforded the language choice.  To wit:

  • Language choice appears as the fifth question in the instrument, begging the question as to how Spanish-dominant respondents would have continued in the poll.
  • Ayers/McHenry do not report whether they used exclusively bi-lingual reviewers or, alternatively, call-backs to secure Spanish speaking respondents.
  • Though only 2% of Florida respondents “opted” to take the survey in Spanish, later results suggest that 33% of their Florida respondents speak mostly or only Spanish in the home, again suggesting a significant push to English completion.

The result is an inherent conservative bias in all results that is driven by the characteristics of their resulting sample, which is too assimilated, too middle class, and too educated to be representative of the underlying population.  To illustrate:

Generation: In 2006, the Latino National Survey (the largest academic survey of the Latino population of the US ever conducted) estimated that foreign-born first generation respondents were 73.5% of Florida’s citizen registered population, 21.2% of Colorado’s registered citizens and 16.3% of New Mexico’s.  By contrast, the Ayers/McHenry poll had only 59% of Florida’s respondents as foreign born, 10% in Colorado, and 7% in New Mexico.  For all three states, foreign born first generation respondents are underrepresented in their data, by 9.3% to 14.2%.

Income:  The Current Population Survey estimates that 47.6% of Latino registered voters have household incomes above $50k, while the Ayers/McHenry poll finds 54% in Colorado and 56% in Florida earned more than $50k.

Education: More telling is the education skew. The percent of respondents with a college education is 40% each in Colorado and Florida and 61% in Florida.  According to the Census Bureau’s Current Population Survey (November 2010 supplement for registration and voting), only 25.1% of all registered Latino voters nationwide have completed college.

Our conclusion is that the Ayers McHenry data demonstrates a significant conservative bias by over-interviewing later generation, more assimilated Latinos and under-representing more working class voters.

Even With the Conservative Tilt, the Numbers Look Terrible for the GOP

The bias we suspect suggests that the results reported in the Ayers/McHenry poll shine a more favorable light on Latino-Republican relations.  Even taking the data at face value, the news is, frankly, terrible for the GOP.  If, as we suggest these results are actually tilted toward the GOP, then these poll results are disastrous news for Republicans.

Ayers/McHenry report dissatisfaction with President Obama and, on this point, we agree.  In all three states, voters are disappointed with the president.  Majorities of all three find the president did not deliver on his promises to Hispanics, and that he is a weaker president than they had expected. This trend has been well documented across our recent surveys.

In light of that result, the Democratic leanings of these registered voters—EVEN IN FLORIDA—are remarkable.

  • Obama polls ahead of an un-named Republican in all three states, by 10 percentage points in Florida, and between 20 and 30 percentage points in Colorado and New Mexico.
  • Majorities in all three states (margins of 5% in FL, 20% in NM, and 24% in CO) believe the president deserves to be reelected.
  • When asked their summary evaluations of parties, organizations, and elected officials, Republicans and their identifiers do not do well.
  • In Florida, by far the most conservative Latino state, Obama’s Favorable/Unfavorable rating is 55/37.  This +18 favorability rating should be compared to:
    • Governor Rick Scott. -19.
    • Republicans in Congress, -6
    • Tea Party, -17
    • Republican Senator Marco Rubio (+23) edges the President in net favorability, but his total favorability (51%) is actually less than Obama’s (55%).
    • Democrats in Congress, +12
    • Senator Bill Nelson (D), +5
  • In Colorado, the President and Democratic officials are widely popular while the GOP brand is VERY weak:
    • President Obama +27 favorable to unfavorable.
    • Republicans in Congress -33
    • Tea Party -32Democrats in Congress +18
    • Democrats in Congress + 18
    • Gov. John Hickenlooper (D) +52
    • Sen. Mark Udall (D) +27
    • Sen. Michael Bennet (D) +32
  • In New Mexico, the pattern is even stronger.  Gov. Susana Martinez favorability is the only GOP bright spot, and she significantly under-performs President Obama and both Democratic Senators:
    • President Obama +33
    • Republicans in Congress -26
    • Tea Party -30
    • Gov. Susana Martinez (R) +15
    • Democrats in Congress +21
    • Sen. Jeff Bingaman (D) +55
    • Sen. Tom Udall (D) +49
  • The net unfavorability of the Tea Party could prove to be a considerable liability to Republicans if nominees are associated with that brand.

The central problem for Republicans remains one that Latino Decisions highlighted some weeks ago, which is that Latinos are liberals and largely do not agree with Republicans on major issues.  Take for example the critical issue of the current economic crisis and Hispanic unemployment.  By margins between 12 and 29 points, Latino voters in all three states prefer government action, spending and investment to more conservative policy ideas—like reducing regulation and cutting expenditures—as the best policy for economic recovery.

And on immigration, Democratic policy preferences like the Dream Act and comprehensive reform with a path to citizenship are strongly preferred to policy positions most often offered by GOP elected officials.  Republicans, meanwhile, receive the lion’s share of the blame for inaction on immigration. Therefore, whether you believe the trends identified by the Resurgent Republic, we believe that there are far more challenges than opportunities for the GOP among Hispanic voters.

Gary Segura is Political Director, Latino Decisions, and Professor of Political Science at Stanford University.


, , , ,