Dick Morris, cut loose from Fox News after years of embarrassingly wrong political analysis, is still gainfully employed by The Hill, which continues to publish his weekly column. The very public battering Morris took following the 2012 election has done nothing to improve the quality of his commentary, and his most recent Hill column demonstrates as much. Headlined “Latinos could be GOP allies,” the column is based on a Republican-funded push poll conducted by a pollster whose work in the 2012 Mexican presidential election distinguished itself with near-Dick Morris levels of inaccuracy.

“A new poll taken by Mexico's leading public opinion researcher shows that U.S. citizens of Latino descent are potentially strong allies of the Republican Party,” writes Morris. The poll, as Morris acknowledges, was “organized and funded by John Jordan of Jordan Winery, a prominent Republican donor,” but that's really the least of its problems. The questions asked, as quoted by Morris, are so over-the-top and one-sided that it's no real surprise the results are so favorable to the GOP:

By 59-34 percent, U.S. Latinos agreed that “Democrats are closer to the leaders we had in Latin America, always giving handouts to get votes. If we let them have their way, we will end up being like the countries our families came from, not like the America of great opportunities we all came to.”

By 78-16 percent, U.S. Latinos agreed that Latino immigrants must "not go the way some have gone into high unemployment, crime, drugs, and welfare.

They must be more like the hard working immigrants who came here and worked their way up without depending on the government." More important, when asked which party most shares this sentiment, they chose the Republicans, by a margin of 45-29 percent.

[...]

By 47-31, Latinos agree that Republicans would do more to “strengthen churches so they can help the poor and teach values of faith and family.” By 89-8, they think that “too many people depend on the government and its handouts. That way of thinking is very bad and leads to lifetimes of unemployment, poverty, and crime.” And, by 45-37, they believe the Republican Party is more likely to share their view than Democrats are.

The poll's methodology does little to enhance its credibility. It combined telephone and in-person interviews conducted over the course of an entire month, January 15 to February 15. What's the problem with having a poll in the field that long? Something big could happen that might alter the opinions of the population being sampled -- like, say, the president of the United States laying out an immigration reform plan.

And what of the pollster, Rafael Giménez? Formerly the public opinion coordinator for Mexican president Felipe Calderón, Giménez signed on as the pollster for the 2012 campaign for Calderón's would-be successor, Josefina Vázquez Mota. (Morris also did campaign work for Calderón.) Leading up to the July 1 election (an election tainted by allegations of vote-buying and coercion), most public polling showed Enrique Peña Nieto of the Institutional Revolutionary Party (PRI) with a huge lead over Vázquez Mota -- except for Giménez's polls.  

In late May, as the majority of polls put Vázquez Mota in third place behind Peña Nieto and Andrés Manuel López Obrador, Giménez released a survey showing his candidate just four points back. In mid-June, Giménez had Vázquez Mota seven points back, and tied with López Obrador. His subsequent poll, released a few days later, showed Vázquez Mota moving past López Obrador and gaining on the frontrunner.

Less than two weeks later on election day, Vázquez Mota finished a distant third, 13 points behind Peña Nieto, and six points behind López Obrador.

In a pre-election interview (in Spanish) with the newspaper 24 Horas, Giménez defended his own polling, and said the reason for the discrepancy was that the polling industry is skewed and that other pollsters are secretly paid by the campaigns. Giménez insisted that his candidate was surging, while Peña Nieto was collapsing:

The election is not decided. There will be a choice, and I'm not saying who's going to win, this candidate or the other, but we have a competitive race and there's a group of pollsters whose job it is to say over and over that it's already decided. They're the pollsters who work for politicians but don't disclose it.

[...]

There's an interesting history here. In previous elections, 1994, 2000, and 2006, the PRI candidate collapsed by six to ten points in the final month. We've already seen it, this isn't a prediction. When the people are not informed, the PRI dominates, but when the people are clued in to what's going on, that's when the PRI falls. This isn't a campaign thing. It's the people.

That's at least not as bad as predicting a “landslide” for his doomed candidate, as Morris so memorably prognosticated for Mitt Romney, but the two in tandem don't exactly scream “credibility,” and their presence in the pages of the Hill doesn't redound to the paper's benefit.