Give any Florida strategist with statewide experience the following data points: by 7:15 p.m., the Democratic candidate has a 10-point lead in Hillsborough, a 100K vote lead in Orange, a 200K vote lead in both Dade and Broward early voting, and is ahead in Duval, and everyone would think the same thing: that Democratic candidate is going to win. Certainly, that is what I thought, and what everyone, R and D, who texted me around that time thought too.
Back in October, I had looked at several different models. Most of them played out with a narrow Clinton win, one of them came back a tie (not in percentages — an actual raw vote tie), and in one of them, where I assumed in most counties that Trump would earn the higher of Romney or Bush 04 vote share, and in that one, Trump won by a point. I sent it to a few friends on both sides, who generally dismissed it. Going into Election Day, pretty much everything was lining up with one of the models that had her headed to about 1.5-2-point win.
I have a plan every election night: check Pasco early vote, then hit refresh until Hillsborough, Pinellas, Duval, Orange, Dade and Broward report, followed by a swing through I-4 suburban and exurbancounties. Sure, the initial Pasco and Pinellas numbers didn’t look too good, but they looked survivable, especially considering pretty much everything else was at or above my target. Then I went and looked at Volusia … Hernando … Brevard … Sarasota … Polk … then back to Pasco. The last of my models was more than playing out.
I slammed down the rest of my beer, and called a buddy in Brooklyn to report the bad news. It was done. CNN could have called it at 8:00 EST, she wasn’t winning Florida. In fact, looking back at my texts, I told a guy at CNN around 8:15 EST that it was done.
Despite my optimism going into Election Day, in my gut, I knew this could happen. As many folks heard me say over the last few years, while I am a big believer — and still am — that demographic trends work in the Democratic Party’s favor, all of this hinges on the Democratic candidate maintaining a reasonable floor with white voters. Frankly, it was a big part of why I was a big proponent of the vice president running. As I told CNN’s “The Lead” in late August 2015 about Biden: “I live in the swing state of Florida. If you look at the way Democrats have struggled with working class, working white voters primarily … he gives us a chance to talk to some voters in the general election that we’ve struggled with the last few cycles.”
President Obama had some reach with these voters, or at least enough for us to win. In 2008, we knew we had to hit 40 with whites, in 2012, we needed to get close to it. For Secretary Clinton, it meant maintaining President Obama’s numbers with whites from 2012. As you will see in a few minutes, she clearly didn’t — not only here, but throughout the country.
So, let’s start with a couple of Florida factoids:
*2016 marked the fourth straight statewide election (two Governors, two Presidentials), where the victor’s margin of victory was roughly a point.
*And just to drive home the point of Florida’s competitiveness — when you go back to 1992, the year where Florida became a true battleground state, there have been more than 50 million votes cast for President, and Republicans and Democrats are separated by 12,000 votes. No, that isn’t a typo — 12,000 votes, or right at 0.02 percent.
*Trump set the new high water mark for Republican vote share in 40 of Florida’s 67 counties.
So, what happened?
I often will describe Florida as a scale. Take the GOP markets (North Florida markets + Fort Myers) and in a neutral year, it will balance out the Dem markets (Miami and West Palm), and more or less, the race balances of the fulcrum of I-4. Because of the Democratic trends in Miami-Dade, the math has changed a bit: Democrats can now count on bigger margins out of their markets than the GOP can out of theirs, and thus can still win even if they lose I-4 by a little bit. This was the Obama 2012 path: the president carried a margin of about 550K votes out of his base markets, Romney was about 410K out of his, and even though Romney narrowly carried both I-4 markets, it wasn’t enough.
Which is a good way to frame the “Things that didn’t cost Hillary Florida” section:
Base turnout: Both Broward and Dade county had higher turnout rates, and the Miami media market had a higher margin for Clinton than Obama. And even with Palm Beach coming in a little short, she won her two base markets by about 75K more votes than Obama 2012, and won a slightly higher share of the vote. Broward and Dade alone combines for a 580K vote margin, and honestly, I think around 600K is pretty close to maxing out.
True, Trump did win the “I-10 corridor” by more votes than Romney, but it wasn’t significant. His 345K vote margin as slightly better than Romney’s 308K, and pretty much in line with Bush 04’s 338K North Florida vote majority. And frankly, Clinton succeeded in the major North Florida objective: keep #Duuuval County close. Trump’s 6,000 vote plurality in Duval County was the best Democratic performance in a presidential election since Carter won Duval in 1976.
It is true that Hispanics underperformed out west, but here in Florida, she did considerably better than Obama in the exit polls — polls that are reflective in the record margins she posted in the heavily Hispanic areas of Miami-Dade, Broward, Orange and Osceola.
This was the GOP talking point during early vote: SW Florida was blowing up for Trump. And they were right, it did. But SW Florida typically has exceptionally high turnout, and high GOP margins, and in the end, Trump’s total was only about 40K votes bigger than Romney.
In fact, if you add up the 8 “partisan” markets, which make up 55 percent of the statewide vote, the 2016 election was basically a repeat of 2012. Trump’s margin was less than 2,000 votes better than Romney.
It was rural Florida: Trump did very well in rural Florida, but so did Romney. If you take all the counties with less than 250,000 residents, he increased Romney’s vote share by 125,000 votes — enough to make up the Obama 2012 margin — except, Clinton increased Obama’s margin in the counties with more than 750,000 residents by over 100,000 votes. In other words, rural and suburban cancel each other out. What doesn’t cancel out — midsize suburban/exurban counties, places with 250,000-750,000 residents — Trump won them by 200,000 more votes than Romney
One more reason HRC ‘cannibalized’ her vote early, in other words, had all the typical Democrats vote early, and lost because there were just simply that many more Republicans left to vote. Here is why this one is tricky.
HRC ‘cannibalized’ her vote early, in other words, had all the typical Democrats vote early, and lost because there were just simply that many more Republicans left to vote. Here is why this one is tricky.
First, Republicans have a lot more “reliable voters” in that, they have fewer voters that drop-off in the midterm elections. Democrats have more “potential voters” — in other words, unreliable or first-time voters. During early voting, GOP had over 200K more “three of three” voters — in other words, people who voted in 2014, 2012, and 2010 who voted early than Democrats, but the Dems had a lot more infrequent voters. And yes, the Dems had more “2012 voters” who voted early, but they also just had more 2012 voters.
Going into Election Day, GOP still had more than 100K “three of three” voters to vote, which alone wasn’t enough to get him to the kind of win he had. However, if you looked at just people who voted in 2012, the GOP edge was just 40K. In other words, had the 2012 voters all voted, the Dem early voting margin would have remained. We don’t yet know who exactly voted on Election Day, but what we do know is the GOP really surged, and Dems didn’t.
In fact, in 10 of the 11 counties where Trump most increased the vote margins from Romney, his vote share (not margin) was at least 6.3 percent higher on Election Day than during early voting — and in six of the 11, the increase was at least 8.2 percent. For example, Trump won 53.8 percent of the Polk County early vote but won 62.6 percent of the Election Day vote — an increase in his share of 8.8 percent. In other words, in some of these counties, Trump was winning Election Day by 15 points more than he won early voting.
And this didn’t just happen in counties where Trump won. Even base Democratic counties saw this Trump surge. Take Broward County, where Trump won less than 30 percent of the early votes, he won over 40 percent on Election Day, or Orange County, where she won early voting by more than 30 points and racked up an almost 120K vote lead, only to watch Trump cut her Election Day only margin to 17K votes. In my last memo, I described what I thought Trump’s Election Day challenge was in golf terms — a 250 yard shot over water. Turns out, he did have that shot. Simply, he crushed her on Election Day.
So, where did he beat her? Simple: I-4, and more specifically, the 15 counties that make up suburban and exurban I-4.
Quick recap: The I-4 corridor is roughly defined as the Tampa and Orlando media markets. If you are a Democrat, win here, and you win. If you are a Republican, win big here, and you win. Given that the rest of the state in 2016 generally looked like 2012, Trump needed to win big here.
But that wasn’t necessarily easy. The urban core in the Orlando market (Orange, Osceola and Seminole counties), is getting more Democratic quickly. In fact, in these three counties alone, Hillary Clinton extended President Obama’s 2012 margins by over 65,000 votes. So, not only does Trump must win the I-4 markets by 75,000 votes more than Romney did in 2012 just to win, he needs to find 65,000 more to make up for urban Orlando.
Well, he did, and more. Trump won the I-4 markets by more than 250K votes. Where Romney won the two-party vote share on I-4 by 2 points, Trump won it by 6 — including winning the Tampa market by 9 points.
But it was even more granular than this. If you break up the markets into two buckets: urban counties (Hillsborough, Pinellas, Orange, Osceola and Seminole), and nonurban counties, the Trump path to victory — and the challenge for Democrats, becomes even clearer.
Despite losing Pinellas County — and Trump’s significant gains there, Hillary Clinton won “urban I-4” by some 200K votes, which was more than Obama in 2008 or Obama in 2012. These counties account for about 48 percent of the votes on the I-4 corridor.
In the other 15, which make up the other 52 percent the region’s votes, Donald Trump won by 450K votes. By comparison, Romney won these counties by 220K votes, and McCain by 130K. In other words, pretty much the entire rest of the state’s election balanced out just like 2012, except one glaring place: suburban/exurban I-4. If you look back at 2004, you will see a similar dynamic.
Here are a few examples:
2008: McCain +7,687
2012: Romney +14,164
2016: Trump +51,899
2008: Obama +13,857
2012: Romney +2,742
2016: Trump: +33,970
2008: McCain +3,135
2012: Romney +7,108
2016: Trump: 26,860
I could go on like this for a while
Overall, Trump won the Orlando market by slightly more than Romney, which is pretty remarkable given Clinton’s strength in the core of Orlando. The Tampa market was solidly Trump. Winning the two-party vote share by 9 points. The rule of Tampa picking Presidents was once again true.
What is interesting is this is also the place where we saw the closest thing to a GOP turnout surge. Of these 15 counties, all but three of them saw turnout rates above 2012, with most seeing their turnout rates up 3-5 points. While these counties are economically entirely different, they are almost universally less diverse than the state at-large. We won’t know exactly who voted on Election Day for a few more weeks, but I would bet we will see some increase in infrequent white voters of all parties to help drive those margins.
Overall, turnout was a bit all over the place this year. The I-10 markets were a smaller share of the vote than 2012, and Orlando was much higher. But within markets, you can see the exurban/suburban thing play out. That being said, Democrats can’t blame this on turnout.
I also think there is an element here of Clinton losing the turnout fight in these places. These were the communities that were not getting a ton of field support (note, I didn’t say none), but were places that Americans for Prosperity were heavily invested in behalf of Rubio. I’ve worried for some time that the “Trump has no ground game” narrative could slowly seep toward complacency, and we might have seen the proof of this in these areas. I wrote about this in a piece on May, when I suggested Trump could win the same way Scott won. Well, it happened.
So what comes next? Well, I will write more on that subject coming soon, but for some of us old guys, we will recognize the 2016 map as very similar to the 2004 map. In the two cycles that followed, Democrats won two statewide races, plus the presidency, and picked up numerous seats in the Congress and Legislature? How? By reaching back into these communities and restarting the conversation. In Florida, the basic rule winning is managing margins, particularly in suburban and exurban I-4. In 04, Bush did it and won. In 08 and 12, Obama won that battle. In 16, Trump did.
And again, this isn’t just a Florida deal — what happened here isn’t isolated. But I will make this one point — one I’ve made a lot over the last few years: if Democrats in Florida can win around 40 percent of the white vote — which is less than what Obama won in 2008, they will win almost every statewide race going forward. Demographics can be destiny — but it isn’t automatically.
Lastly, to the organizers on both sides — stay in the fight. If you were for Trump, go be a part of the solution. President Obama told his 2008 organizers to go make their own solutions — you should too. For the Clinton organizers, get up off the mat. There are more fights ahead and more chances to contribute.