It's not hard to understand why many believe Rubio would offer Romney a boost in the Hispanic vote. In the country's entire history, no major party ticket has ever included an Hispanic candidate as the presidential or vice presidential nominee. When Barack Obama defeated John McCain in 2008, McCain got only about 33% of the Hispanic vote; George W. Bush had won about 44% in 2004. Thus, there's plenty of room for improvement--a Romney/Rubio ticket would only need to get back to 44% in order to give Obama a major headache. And Romney himself is not particularly repulsive to Hispanic voters: Romney won the Hispanic vote overwhelmingly during the 2012 Republican primary.
However, there are many skeptics. Among Democratic analysts, the conventional wisdom is that Rubio would have a negligible effect on the Hispanic vote. Their argument is essentially as follows: While Rubio is Hispanic, he is Cuban-American rather than Mexican-American. Mexican-Americans comprise roughly two-thirds of America's Hispanic population, while Cuban-Americans account for about 4%. Rubio would influence the Cuban-American vote, but they already lean Republican. According to Democratic analysts, Mexican-Americans (and presumably other Hispanics) have a negative attitude toward Cuban-Americans and do not consider that group "one of them." They argue further that Hispanics are not monolithic and would not vote based on ethnicity anyway.
The skeptics' analysis, while a source of comfort to many Democrats, contains a number of major flaws. First, there is a basic misunderstanding of what Rubio would need to do in order to help Romney win the election. Rubio would not need to replicate Obama's 2008 feat of capturing more than 90% of his ethnic group's vote. As mentioned, Bush got 44% of the Hispanic vote, which is roughly the point at which Obama's coalition would begin to collapse. Rubio would simply need to peel off a small chunk of Obama's Hispanic vote share--including many who voted Republican in prior elections.
Next is the question of whether non-Cuban Hispanics would be influenced by Rubio's presence on the ticket. An analogy might be drawn to what occurred during the Democratic primary in 2008. Initially, polls showed black voters supporting Hillary Clinton over Obama by a wide margin. Analysts--especially those supporting Clinton--offered many reasons why black voters would prefer Clinton to Obama. They claimed that while Obama was black, he was not considered "black enough." Obama was biracial; he was raised by his white grandparents; he grew up in Hawaii and Indonesia rather than a big city; he had a privileged background; his father was an African immigrant rather than an African-American; he was not descended from slaves, etc. At that time, analysts also remarked upon the fact that the black vote is not monolithic, and that black voters would not vote based solely on ethnicity.
Polls continued to show Obama trailing Hillary Clinton; the Clinton family was always well-received by black voters. But after Obama won Iowa, it became clear that Obama had a real chance to become the first black president. At that point, everything changed. The black vote unified behind Obama, who frequently defeated Clinton among that ethnic group by margins of 10-to-1. The Clintons even found themselves repeatedly accused of racism when they criticized Obama. Evidently, there was no longer a perception that Obama was not "black enough" to attract black voters. Obama may not have been born to African-American parents, but he was the "most" black candidate who had ever made a serious run for the presidency.
Similarly, Rubio would be the "most Hispanic" candidate to appear on a presidential ticket. He would not need to cause a mass defection of Hispanic Democrats the way Obama caused nearly all of the black Clinton supporters to switch sides; Rubio would merely need to regain the Republicans and some Independents who voted for Obama.
While voters may not vote solely on ethnicity, it's difficult to argue that the first Hispanic VP wouldn't draw the attention of Hispanic voters and cause them to consider the Republican ticket on a more serious level. In addition, for those who avoid voting Republican for fear of anti-Hispanic bias in that party, the mere presence of an Hispanic on the ticket would have a real impact.
If Rubio were on the ticket, we might expect a parallel to Obama's experience in the Democratic primary to play out as follows: Rubio is announced as the VP nominee. Threatened, Democrats attack the choice as "cynical." Early polls show little movement toward Romney among Hispanics; Democrats declare that Rubio has failed to have an impact. Democrats then hone in, attacking Rubio's lack of experience. Over time, as the possibility of an Hispanic vice president becomes more real, and as Rubio makes his case, things begin to turn. Hispanic voters slowly shift toward the Republican ticket. Failing to notice this, the Democratic attack machine overreaches--they are unable to resist comparing Rubio to Sarah Palin. This backfires.
Rubio is viewed as intelligent and passionate about the American dream. By claiming Rubio is unintelligent, Hispanic voters feel the Democrats have treated Rubio unfairly. They don't agree with Rubio on everything, but the Palin comparisons are insulting and suspicious. Dangerously, signs of anti-Hispanic prejudice leak among some less-disciplined Democratic surrogates. As the VP debate approaches, expectations for Joe Biden are inexplicably pumped up by optimistic Democrats who temporarily forget his history of gaffes. Rubio crushes Biden in the debate. Biden responds with increasing condescension toward Rubio, and it doesn't have the desired effect. Only after election day do Democrats realize how large a share of the Hispanic vote the Romney/Rubio ticket has managed to get.