Reshaping the Electorate

I've been really blessed in my career.  For a kid from a small town in Florida, by way of a small town in Illinois, I've enjoyed several "how the heck did that just happen" moments.  And without question, one of my more favorite memories was working the Presidential debate spin room at Lynn University, "on background" to make the case that Florida would go for President Obama, even when a number of misinformed pundits were suggesting the President wasn't even really competing there.

My argument was simple.  We had a superior organization, but arguably just as important, the state's demographic trends meant the electorate would be made up of even more Hispanics, as well as African Americans and Caribbean Americans than 2008.  This was definitely a sentiment that was outside the conventional wisdom, which somehow fell to the belief that the proportion of the electorate made up by ethnic minorities would be lower in 2012 than 2008.  In fact, this conventional wisdom led to some downright crazy statements about Florida including the traditional "they aren't really trying here" to the absurd: "Romney will win Florida by seven."  

In the end, demographics shaped the outcome.  Ethnic minorities made up as much as 33% of 2012 voters, at least 3, if not 4 points higher than 2008 and the rest is history: President Obama won a narrow victory here, becoming the first Democrat since FDR to carry the state in successive elections.  

While the turnout models in non-Presidential years in Florida are always different, ethnic minorities in 2014 will make up a larger share of the electorate than 2010.  But how much more?  And is it enough to alone change the outcome?  That is the purpose of this piece.

Let's start by looking back at 2010.  In what was arguably the best year for Republican year in Florida history, exit polls show the electorate broke down like this: 74 percent white, 11 percent black (both African American and Caribbean American), and 12 percent Hispanic.   Going into that election, voter registration lined up this way:  69 percent white, 13 percent black and 12.5 percent Hispanic.   All of those things added up to a 61,000 vote victory by Governor Scott.  In other words, the electorate was significantly more white than voter registration.

Before we go further, you might be asking yourself -- how is it that the electorate was made up of more ethnic minorities in 2012 than their proportion of registered voters, when participation among black and Hispanic voters is traditionally lower?  Two things:  turnout among these groups was higher in 2012 -- and in the case of Hispanics, there is a segment of the population who will register to vote as white, but self-ID to a pollster as Hispanic.  This is part of why polling this subgroup in Florida is tricky.

Back to the data -- so what has happened with these population trends since then?  For the sake of this exercise, and in an attempt to compare apples to apples, I will look at the data going back to the close of the registration books for the 2010 primary, roughly the 1st of August 2010.  

Here are a couple of interesting points:

• Despite the fact that some 2 million people have registered to vote since August 2010, the actual number of active voters is roughly 500,000 more than 2008.  This is due to mortality and non-voters being moved to 'inactive' status.
• Since 2010, 20 percent of all new registered voters are self-identified Hispanic, 15 percent are black (African American or Caribbean) and another 9 percent are other.  This is likely a mix of people, including the state's fast growing Asian population and some mixed ethnicities.  The remaining 56 percent are self-identified white -- though keep in mind, some Hispanics, particularly those whose families have been here multiple generations, self-identify as white.
• This leads to an electorate that today is roughly 66 percent white, 13.5 percent black and 14 percent Hispanic and with a larger proportion of Asian and other ethic populations. 

So how are these new voters registering?

• 97 percent of black voters (African American/Caribbean American) are registering either Democratic (73%) or NPA (24%).  Only 3 percent of new black voters are registering GOP.
• Only 22 percent of new white voters registered Democratic.  The plurality of new white voters registered NPA (40%), with the GOP close behind (38%).
• On flip side, only 16 percent of new Hispanic voters registered Republican.  Like whites, the plurality of Hispanics registered NPA (45%), with 39 percent joining the Democratic Party.

As a result, the ethnic make up of the political parties looks like this.  The below percentages are the share of each group within the political party:

                Black Hispanic White

Democratic 28.4% 14.0% 52.4%

Republican   1.3% 11.2%  83.7%

NPA           7.5%  18.9%  63.4%

Total         13.4% 14.2%  66.4%

The other way to look at it is partisan registration by ethnicity:

                       Democratic Republican NPA

White (66.4%)     30.6%      45.0%      24.4%

Hispanic (14.2%)  38.0%      28.1%      33.9%

Black (13.4%)      82.1%      3.5%       14.6%

Other (6%)         34.5%      22.0%      43.5%

It breaks down this way: political reporters, observers, pundits, players, etc., are all looking for that silver bullet to call the Governor's race.  Will it be Scott's money advantage, will medical marijuana turn out voters, will the Affordable Care Act rally Republicans -- or Democrats, will base Democrats stay home, will Tea Partiers show up, will the Libertarian candidate take votes from Scott, Crist or just be a protest candidate,  will the Jaguars take a QB in the first round, how far can Jameis Winston throw a crab leg, it Tony Stewart continues to struggle, what happens to turnout from transplanted Indiana race fans, what does the LG pick mean, etc., etc., etc.  Story after story will be written about how each of these is the absolute critical piece to one candidate's success or failure. 

Truth is from my perspective, all of it and none of it matter. Florida is a marginal state.  We aren't a marginal state because we are a swing state, in the way of an Iowa, which has a lot of swing voters.   No, we are a swing state because we are this amazing collection of subgroups, that when added together, add up to be a competitive electorate.  The number of people in this state who are true swing voters is not huge -- though for people who argue that Democrats win by simply turning out their vote, I point you to 2004.  Florida is both a persuasion and a turnout state.

So that is a long way of saying, yes, the demographics matter. Let's say Charlie Crist takes Scott's marginal win among Hispanics in 2010 and makes it a 53-46 Crist win.  Should Hispanics make up 13% of the electorate, that change from 2010 to 2014 will add up to essentially another point on election day for Crist. For blacks, if Crist can get turnout from 11 to 12 percent of all voters, again, that means nearly an entire point accrues to Crist's total.  Bluntly, all of these stats simply mean that the Democrat needs a smaller percentage of the white vote to win.  

And remember, Scott won by barely over a point, so it is all about margins. The changing demographics simply mean that window for him is just that much narrower. 

Comments

[M.J. Wilson]

Very interesting analysis of the changing demographics of Florida. Very interesting.