Public polling is such an imperfect science. Some polling firms such as Gallup, Pew Research, Rasmussen, and Zogby do a good job of gauging the sentiment of the people. Others, not so much. The best you can do is look at a lot of polls from different organizations at different times, and identify some patterns and come up with some basic conclusions. The four major presidential polls conducted in Honduras during January-May by CID-Gallup, Le Vote, and Encuestadora Paradigma all showed Xiomara Castro de Zelaya of the Libre Party ahead by a clear margin. All but one of the polls had Liberal Mauricio Villeda last by a significant margin. There's one pattern for you. Two of the polls had Nationalist Juan Orlando Hernández in second place, while the other two had Salvador Nasralla of the Anti-Corruption Party in second. There's another pattern.
The fact that Mrs. Zelaya consistently polls high, while Mr. Villeda consistently polls low suggests that Mr. Villeda may have a lot of work to do to get his message across and improve his image. That's one basic conclusion. The fact that Mr. Hernández and Mr. Nasralla have each come in second twice suggests that the real battle right now is for second place. That's another conclusion.
Toward the end of May, Channel 10 in Honduras conducted a presidential poll. The results had Mr. Villeda on top with 53.13%, followed by Romeo Vásquez of the Patriotic Alliance Party with 25%, Mr. Hernández with 15.63%, Mr. Nasralla with 3.13%, and Mrs. Zelaya with 3.13%. This poll is widely at odds with the ones by CID-Gallup, Le Vote, and Encuestadora Paradigma, so it raises questions about methodology. But it should still be taken into account.
A few days ago, the Tigo cell phone company polled 2,361 of its customers about the presidential race. It asked the question, "At this moment, which of the candidates do you believe has the strongest appeal to the Honduran electorate?" The results showed Mr. Villeda coming out ahead with 54.13%, followed by Mr. Hernández with 24.90%, Mr. Nasralla with 12.33%, Mr. Vásquez with 4.45%, and Mrs. Zelaya with 4.19%. Interesting, in that Mr. Villeda again polled above 50%. Note, however, that we don't know the methodology used. For example, did Tigo allow its customers to vote multiple times, and were those voting charged a fee each time they voted? On both questions, our assumption would be yes. Does it matter?
Honduras Weekly is now conducting its own presidential poll. We are polling as many different Honduran online social networking sites as possible, as well as our regular audience. We have made a conscious effort to poll sites that have a wide variety of political leans. We began polling on May 29, and we will continue polling until election day. A total of 137 people have been polled through June 4. Mr. Villeda currently leads with 36.50%, followed by Mr. Nasralla with 24.82%, Mrs. Zelaya with 21.17%, Mr. Hernández with 13.14%, Andrés Pavón of the UD-FAPER Party with 2.19%, Jorge Aguilar of the PINU Party with 1.46%, and Mr. Vásquez with 0.73%.
It is not surprising that our presidential poll has met with some initial criticism. Some of the criticism has been subtle, and some of it has been fairly blunt. The subtle criticism goes like this: "So you are polling only those with access to a computer? That leaves out a lot of Hondurans. And the poll allows non-Hondurans to vote.
The underlying concern seems to be that our poll is not sufficiently broad or expansive. That's true, but it's true for most polls. All of the presidential polls that have been conducted in Honduras leave out a lot of Hondurans. Most polls will poll no more than a few hundred or a few thousand people. The same is true in the United States. Polls are designed to take a relatively small sampling of a population. Honduras has about 8.5 million people, so... yeah, a lot of Hondurans are going to be left out.
Our poll began on May 29, and we intend to continue it for several months. We hope to eventually poll at least 1,000 people.
And yes, our poll allows non-Hondurans to vote. This is clearly a flaw in our methodology. Unfortunately, the online polling system -- Survey Monkey -- we are using has no way to distinguish Hondurans from non-Hondurans. This concerns us a little, not a lot. We are making a huge effort to poll Honduran social networking sites, and so we believe a vast majority of respondents will be Honduran.
The blunt criticism we have encountered has gone like this: "127 people polled? Is this a joke? The company that conducted this should be ashamed of making public a poll with only 127 interviews..." Or... "How unprofessional. Who are these mercenaries who did the research?" Or finally... "127 people consulted, it's not a believable and credible result. There is no doubt that they had already had everything fixed."
No, there's nothing "fixed" about it. It's an open online poll, which anyone is free to respond to. So far, we have a very limited number of respondents (137, at last count), but remember that we just started polling a few days ago, and we plan to keep it going for a few months. The poll results will depend solely on how many people opt to get on their computers, click on Poll... and vote. Really, very forthright.