The late Washington Post prize-winning reporter David Broder shook up the political world 35 years ago with his book Changing of the Guard: Power and Leadership in America; in it, he correctly predicted the demise of the long-entrenched Democratic Party of the South but missed today's Hispanic political growth and effect by a million miles.


His problem in dismissing Hispanics was he did not understand the community, did not see what many others could see and had no knowledge whatsoever of what Hispanics had already accomplished politically in U.S. history.

As brilliant as that book was, as was his Washington Post work, if he were alive today his work would be marginal politically. His view on Hispanics was in a black-and-white context only, then, and would be today.

There are 55 million Hispanics in the United States. Hispanics aren't a race; they can be of any race and have different national origins. Their median national age is 27, while the non-Hispanic population has a median age of 40-plus. While lagging in education for decades, recent graduating classes of 2012 and 2013 have burst through the 80 percent graduation rate. Along with exponential growth in college attendance, Hispanic doctorate candidates reflect that growth and are second only to fellow minority Asian-Americans in numbers.

My alma mater, San Diego State University, reflects this revolution. When I enrolled in 1958, there were 7,000 total students and maybe 50 Hispanics. Today, there are 25,000 total students, of which 5,000 are Hispanic undergraduate and graduate students.

Taking a macro approach when looking at Hispanics works when one looks at total population, demographics, geographic distribution, income levels, etc. Where it breaks down, however, is in political analysis.

When one sees political surveys or exit polls of Hispanics, one important analytical ingredient is missing; to wit: the significant differences between the various Hispanic groups that are differentiated by national groups and economics.

For example, most national exit polls of the 2012 presidential election (President Obama versus Republican nominee Mitt Romney) concluded, according to the media-paid-for exit polls, that Obama's 71 percent of the Hispanic vote squashed Romney's 27 percent. Similar exit poll results were indicated in the 2014 midterm elections.

These numbers are highly misleading because there is no breakdown of the numbers by ethnicity or national grouping and key Republican victory states were not measured at all (Nevada and New Mexico).

Nationally, the Pew Research Center reported in 2012 that of the 55 million Hispanics, 35 million are of Mexican origin (65 percent); 4.7 million are Puerto Rican (9 percent), 2 million are Cubans (3.7 percent), 2 million are Salvadorans (3.7 percent), 1.5 million are Dominicans and 1.1 million are Guatemalans. There are 2.8 million other people of Central and South American origin, as well.

Here is where the political analysis of Hispanics falls apart: Those numbers don't tell the political story.

For example, the most loyal Hispanic voters in the country are Democratic Puerto Ricans and Dominicans; they are centered in New York, New Jersey and Central Florida. Subtract out the Puerto Ricans and Dominicans from the New York and New Jersey votes and both states would go Democratic without Hispanic Democrats. Thus, their vote is nonessential for victory.

Florida is different. Florida Puerto Ricans have offset 50 years of Republican domination of South Florida Hispanic votes. Florida is a toss-up state in presidential elections because of Puerto Ricans. Thus, they are politically important in Florida, as they are not in New York.

On the other hand, in last November's Florida gubernatorial race, over-40 Miami-Dade Cubans voted 70 percent for Republican Gov. Rick Scott, giving him a 60,000 vote statewide victory out of millions cast, by keeping Democratic margins in staunch Democrat Miami-Dade and Broward Counties to minimums.

Out West, Hispanics voted heavily for Republican Gov. Brian Sandoval in Nevada, where Obama won in 2008 and 2012. Sandoval won with 70 percent of the total vote. In New Mexico, Republican Gov. Susana Martinez won reelection in a state Obama carried twice; she won with 56 percent of the vote. Hispanic exit polls in Nevada and New Mexico were not done; thus, we have no measurement as we do elsewhere.

Nationally, Hispanics gave 62 percent of their votes to Democrats and 36 percent to Republicans in the 2014 midterms.

However, several individual state races showed significant Hispanic differences from the national results, with the differences based on Mexican origin. The more Hispanic voters of Mexican origin, the more GOP votes.

In the 2014 Georgia Senate race, for example, 42 percent of Hispanics (mostly Mexican) voted Republican; 47 percent voted Republican for governor. In Texas, mostly Mexican voters gave 47 percent of their vote to Republican Sen. John Cronyn and 44 percent to Republican Greg AbbottGreg AbbottSunday shows preview: Multiple states detect cases of the omicron variant Hillicon Valley —TSA to strengthen rail sector cybersecurity When politics trump workers' health, we know who gets burned MORE for governor.

Analysts, pollsters and scholars need to study the various groups within the overall Hispanic community in the future. If they don't, national exit polls are meaningless, especially when looked at by political parties in developing strategies for appealing to the burgeoning Hispanic vote.

Every single month sees 10,000 new Hispanic voters. Even David Broder would realize how significant 10,000 new voters a month are and will be in the future.

Contreras formerly wrote for Creators Syndicate and the New American News Service of The New York Times.