One of the Republican Party's biggest problems in the 2012 presidential election was their lack of appeal to minorities. The GOP always wins white voters, but in
2012 they only won 6 percent of blacks, 26 percent of Asians, and 27 percent of Hispanics. Compare that to
2004, when President George W. Bush won 11 percent of blacks, 43 percent of Asians, and 44 percent of Hispanics. President Barack Obama's 2008 and 2012 campaigns were successful because they developed a coalition of Hispanics, African-Americans, young people, and single women. The next Democratic nominee wants to repeat this while the next Republican nominee wants the coalition to fall apart.
Some political experts argue that Mitt Romney lost in 2012 because he failed to extend his support to different racial groups. If he picked Marco Rubio, who was talked about as a VP candidate in 2012, then Romney would have improved among Hispanic voters. In the end, he picked Representative Paul Ryan of Wisconsin. Ryan was move conservative than Romney, so he could rally the Republican base. The plan didn't work and Obama won a second term. In fact, there was
a decrease of 6.5 million white voters in 2012 (they didn't win the swing state of Wisconsin either). These voters could have been conservatives who didn't like Romney.
The loss of white voters is the most important demographic change in 2012. The Republicans do have options. They can go after these voters to win in 2016, but they can also increase their appeal among Hispanics and African-Americans. They don't need to win black voters, but a simple increase of 5 percent to return to Bush's numbers will help them a lot. They should also want to get back to the support they had with Hispanics in 2004. I think it is important for the presidential nominee or the VP nominee to be of a different race to broaden the ticket. Sociological voting matters.
RealClearPolitics has now set up a cool
simulation that can help people figure out how many voters the Republican Party needs. I'm going to use this simulation to find a pathway to victory for Republicans in 2016. The only problem with it is that third parties are not a factor (but the number of votes that will go Libertarian, Green, Constitution, or to other minor parties is always unpredictable). I decided to keep turnout as it was in 2012 (64.1 of whites, 66.2 percent of African-Americans, 48 percent of Hispanics, and 49.3 percent of Asians voted in the election).
Same Turnout as 2012
First, I decided to increase the GOP's share of the black vote from 6.1 percent to 9.5 percent, which is roughly the average for what Republican candidates win. Next, I decided to increase the share of Latino votes for Republicans by 10 percent (from 27.6 to 37.6 percent). I also did this with Asians, so the GOP share with that group increased from 31.6 percent to 41.6 percent. Finally I increased the share of white voters just 1 percent (from 60.2 percent to 61.2 percent). I got this:
Those changes shifted four battleground states to the Republican side. These were Florida, Virginia, Ohio, and Colorado. That's 275 electoral votes for the Republican and 263 electoral votes for the Democrat. The Republican won 50.6 percent of the popular vote while the Democrat won 49.4 percent. Remember, third parties are not accounted for.
We can't say that a Republican victory in 2016 is going to look exactly like this. I'm only assuming that it will be close right now, but catastrophes do occur. If Rick Santorum is the Republican nominee, maybe he would like to focus his efforts on his home state of Pennsylvania (a swing state) rather than somewhere else. Marco Rubio and Jeb Bush probably wouldn't have to worry about Florida like other Republicans who aren't from there would, so they can can spend more campaign time in other states. Before he dropped out, Scott Walker probably would have focused on Wisconsin and Iowa if he was the nominee.
Adjusting Turnout
One criticism I predict in receiving is that I did not adjust turnout among the groups. Some political commentators argue that white turnout is going to decline while turnout among the other three groups will increase. In response to that criticism, I decided to run the simulation again and decreased white turnout from 64.1 percent to 62.1 percent. I increased African-American turnout from 66.2 percent to 68.2 percent, increased Hispanic turnout from 48 percent to 50 percent, and increased Asian turnout from 49.3 percent to 51.3 percent. Under this simulation, I kept the Republican share of voters the same as I did with the first simulation. This happened:

Now the Democratic candidate wins with 294 electoral votes to the Republican's 244 electoral votes. It's a nail-biter though, with both candidates winning 50 percent (roughly 65.3 million for the Democrat and 65.2 million for the Republican to be precise). If turnout grows, demographics will hurt the Republican Party unless they extend their message to other racial groups.
Adjusted Turnout with a Republican Victory
The Republicans need more votes to win Ohio and Virginia, which is their path to victory. I increased the black vote to 10 percent, increased the Hispanic vote to 40 percent, and increased the Asian vote to 42 percent. I did nothing with the white vote. Under that simulation, the GOP is back to 275 electoral votes with 50.4 percent of the popular vote. It looks like this:
Reaching these numbers isn't that hard for Republicans. They are less than what Bush won in 2004. In fact, this same simulation can be run if the Republican candidate wins an additional 1 percent of whites while doing nothing with other groups (that means the GOP vote share for whites is 62.2 percent).
Adjusted Turnout with 2004
The last time Republicans won a presidential election was in 2004. I decided to use Bush's numbers in that election with my simulation. I find this simulation very interesting because Bush won less whites than Romney did and it hurts Republican chances today:
The Republicans improve, but they only win in Florida.
From looking at these simulations, it is clear that the GOP does need to expand support among minorities, but white voters still matter. If they improve by 1-2 percent with white voters, if they improve by at least 2 percent with blacks, if they improve by over 10 percent with Hispanics, and if they improve by over 10 percent with Asians, then the Republican presidential candidate will be in good shape.