Monday, September 9, 2013

New July Polling in the Presidential Race

A friend dropped the 5th CESPAD survey of public opinion in Honduras in our in box, this one for July, 2013.  CESPAD is the Centro de Estudios para la Democracia, and came into existence in 2010.  Since then they've been performing polling on Honduran public sentiment.

In this case we're talking about a survey of 1,440 individuals over 18 years of age in a probabilistic, stratified, multi-stage sample between July 21 and July 31 of 2013.  They report margin of error of plus or minus 2.5%.

The survey focused on a number of topics.  This post will report on the sections about the November, 2013 elections; there will be more on other topics to come.

CESPAD found that these people plan to vote.

Overall, 80.3% of those surveyed said they intend to vote. Honduras typically has a voter turnout around 50-55%.

CESPAD asked respondents who planned to vote what party they would vote for in 2013.  Here's the results

     Libre           26.7%
     National      24.1%
     Liberal        16.6%
     PAC            10.2%
     Nobody       12.3%
     No Answer   8.6%


Every other party was less than one percent (Sorry Romeo!).

Over the last two years of polling, CESPAD found that the number of people who would vote for the National Party and Liberal Party has steadily declined. In the current election, the percentage who intend to vote for Libre and Partido Anti-Corrupción (PAC) take up the slack.

The popularity of each party's presidential candidate is similar to, but not the same as, the party popularity:

     Xiomara Castro (Libre)                     28.0%
     Juan Orlando Hernandez (National)   20.7%
     Mauricio Villeda (Liberal)                  13.7%
     Salvador Nasralla  (PAC)                  11.7%
     Undecided                                       17.8%
     Nobody                                             6.8%
     Romeo Vasquez (Alianza)                   0.7%
     Orles Anibal (Christian Democrat)       0.3%
     Andres Pavon (FAPER)                       0.2%
     Jorge Aguilar ( PINU)                         0.1%


Like other polling, CESPAD's numbers indicate that Hondurans could elect a president in November who receives far less than 50% of the vote. This could create serious complications for governing, depending on the composition of the Congress.

CESPAD notes that this is essentially a two-person race between Xiomara Castro and Juan Orlando Hernandez, both of whom have seen rising support over the last six months.  CESPAD concludes that the data suggest a reconfiguration of electoral politics is underway in Honduras, but the trends aren't yet definitive.

More interesting still is where each candidate's support comes from.

Slightly less than half of Liberal Party members surveyed said they would vote for the Party's candidate, Mauricio Villeda.  Xiomara Castro picks up support from 23.1% of them, and Undecided/Nobody a further 16.7%.  Nasralla and Hernandez pick up minor amounts of support here as well.

The National Party is a bit more cohesive, with 58.4% of National Party members saying they'd vote for Juan Orlando Hernandez, the party's candidate, and a further 20.8% Undecided or Nobody.  Villeda, Castro, and Nasralla pick up minor amounts of support among National Party members.

PAC is still more coherent than the National Party, with 81.9% of party members saying they will vote for their party's candidate, Salvador Nasralla, and a further 11% being either undecided or planning to vote for none of the above. Villeda, Hernandez, and Castro each pick up single digit support among PAC members.

Libre appears to be the most cohesive party, with 94% of its members going to vote for Xiomara Castro, the party candidate, and only 3.8% undecided. Villeda, Hernandez, and Vasquez each pick up support from less than one percent of Libre's members.

CESPAD concludes that a real effect of the 2009 coup has been to disrupt the two party system in Honduras.

Party fidelity appears to be the casualty, in line with party fidelity trends in other latin american countries. Only a third of respondents said they would never vote for a party they didn't belong to.  Over 50% said they might vote for the candidate of a party they didn't belong to, and nearly 20% said for sure they would be doing so this time.

The better educated the voter, the more likely they intend to vote for Castro or Nasralla.  The traditional parties do best with those with little or no education.

The take-away from this poll, as in others, is that the real race is between Libre's Xiomara Castro and Juan Orlando Hernandez of the National Party.  Villeda does not seem to be a viable candidate, and he loses a nearly a quarter of his party to Xiomara Castro.  Party loyalty suffered in the National Party as well, and Hernandez has less than 60% of his party behind him.  In order to win the presidential election, he will need stronger support from Nationalists.

Nasralla is running well with a consistent 11-15% of the electorate supporting him-- not enough to be elected, but enough to suggest his platform resonates.

These numbers bear no resemblance to those reported by Paradigma for July. Both polls show the candidates in the same order. The differences in absolute numbers may lie in how questions were asked, and the sampling strategy of each polling group.

Paradigma now reports the results of its August survey.  It shows Libre, the National Party, and the Liberal party numbers up, with Nasralla and none of the above falling:

     None of the Above  27.0%
     Xiomara Castro       22.9%
     J Orlando Hernandez 19.9%
     Salvador Nasralla    10.3%
     Mauricio Villeda       9.4%

Variations of 2-3% from the July poll results are within the margin of error of the poll, so none of these changes are significant.

Neither polling company has been through an election cycle in Honduras yet, so we don't know how they'll fare against actual election results.

All the polls, though, are telling the same story: a fractured electorate, a surging new party (Libre), and a sinking old one (Liberal).

No comments: