I Watched the Democratic Collapse in Florida. I Fear It’s Happening Nationally.
It’s time for real structural changes in the party, not quick fixes. Here are a few.
IT IS TIME TO STOP talking about 2024.
Don’t get me wrong, I have a lot of opinions about the campaign and the years leading up to it. But that isn’t the conversation Democrats need to be having. The real conversation is how my party went from the broadest electoral mandate of the previous twenty-five years, with the biggest majority in the Senate in the previous thirty years, to a shell of itself—a political organization that can hardly be classified as a national entity anymore.
There are a lot of metrics we can cite for this, but this may be the most telling: Of Barack Obama’s 332 electoral votes in 2012, 53 of them came from states that aren’t really in play anymore: Florida, Ohio, and Iowa.
All three states have gone Republican three elections in a row. Prior to 2012, Iowa had gone Democratic in five of the previous six presidential elections, Ohio had gone to Democrats in four of six, and Florida in three of six (though it probably was four out of six had my state known how to count ballots).
That’s not just a canary in a coalmine. It is a massive boulder landing in front of you on the only road home.
Even worse, my party has largely avoided reckoning with how big that boulder is.
We can no longer do that or we will find ourselves in an even worse situation than we do following the 2024 election. Now is not the time for quick fixes. We must make real structural changes.
More money must be spent earlier. It isn’t like the challenges this cycle weren’t obvious. Poll after poll I was involved with showed hardening of opinions on the economy, crime, immigration, and other issues. Democrats needed much earlier investment communicating on all of these issues. We failed to make them. This stood in contrast to 2020, when more outside groups and the party made earlier investments. Unite the Country, the super PAC I ran that cycle, ran ads starting in late April 2020 and communicated with voters all the way to the end of the cycle.
We need more groups involved. Both David Plouffe and Jen O’Malley Dillon made this point in their appearance on the Pod Save America podcast, recapping their perspective from the helm of the Harris campaign. Having more groups doing more work helps the nominee. In both 2016 and 2024, virtually all the super PAC money ran through a single entity, and both times we lost. The 2020 approach, which happened as much by circumstance as by intention, meant more groups were at the table, bringing a diverse set of opinions and skills. The governing theory of the case going into the 2020 general election was the race was all about Trump. What we collectively found through research and sharing ideas was that it was really about introducing and creating comfort with Joe Biden. Without that diversity of thought, I don’t know that we would have gotten to the right spot in 2020. In a race won by fewer than 50,000 votes in three states, it mattered.
We must be smarter about how we use data. Right now, we use data as a crutch. We were addicted to ad-testing, to the point that it drove decision-making more this cycle than the desire or need to tell a story. We overuse analytics to find the most “efficient” ways to communicate with voters, meaning in many cases, we just don’t talk to huge swaths of both our base, or to Republicans. Data also allow campaigns to avoid accountability for decisions—just blame the outcome on following the analytics. Data are vital, but should work for the campaign—not the other way around.
We must deal with the right’s tremendous advantage in delivering content. After 2020, I had a billionaire ask me what I thought would be useful going forward. My advice was to spend a billion dollars building out an ecosystem like the right to deliver information to not only our base but persuadable voters. There was an acknowledgement of the problem, but that was all. I worry that coastal Democrats don’t fully grasp just how much of a disadvantage we face on the news consumption front—especially podcasts and social media—and that to solve it, we need a donor or two willing to invest significant capital.
Ultimately, these tactical changes matter. But none will fix the underlying problems that have only been growing for Democrats. Campaigns are about addition but, increasingly, our coalition is contracting. For example:
The three most Hispanic counties in Florida—Miami-Dade, Hendry, and Osceola—went from giving Hillary Clinton a 324,000-vote margin in 2016 to giving Donald Trump a 133,000-vote margin just eight years later.
In just twelve years, the county of Joe Biden’s birth, Lackawanna County in Pennsylvania, went from giving Obama a margin of almost 30 points to giving Harris a margin of about 3 points.
Within the last fifteen years, Democrats have had senators and House members in states like Alaska, Idaho, Oklahoma, Nebraska, and the Dakotas. Minus the occasional blip in a wave election year we really aren’t even relevant to the political conversation in those states anymore.
None of these things happened overnight, and all of them mattered way more than the Harris campaign not doing Joe Rogan’s show.
The truth is we got here because our brand sucks. We tend to put voters in different buckets—black, Hispanic, young, gay, etc.—and treat these groups like they are more progressive than they really are, and somehow unique from each other. At the same time, we’ve made decisions to stop talking to large chunks of the electorate.
Some of these decisions can’t be fixed. For example, I wish we would have been more intentional about molding the Obama operation into something more permanent for the party to utilize. Coming out of 2008, rather than investing in partisan infrastructure, many in my party—in some cases led by the Obama White House—encouraged the development of “long-term progressive infrastructure,” with the idea that not-for-profit organizations could better address political needs, instead of state parties. Not only has this experiment failed at the core organizing level in states like Florida, but it has encouraged the idea that Democrats are beholden to progressive groups and values. We should have continued something akin to the old Howard Dean plan to invest in all fifty states.
But we have a bigger problem. Sure, we can win elections under the right circumstances, but we no longer have anything remotely close to a long-term winning coalition.
Florida Was a Warning
THERE WERE VISIBLE CRACKS in the Democratic coalition prior to 2024. The challenges my party has with blue-collar white voters first started to appear in the off-cycle elections of 2010 and 2014. But it really showed up in earnest in 2016.
For me, the first sign Clinton was likely to lose the election in 2016 was Volusia County, in my home state of Florida. A multiracial, blue-collar county that is home to Daytona Beach and NASCAR, Volusia was a reliably Democratic county. How reliable? It had gone for Democrats from 1992 to 2008, with Obama carrying it by six points, or about 14,000 votes. Romney narrowly beat us there in 2012, but it remained in the top fifteen best vote-share counties for Democrats in Florida that year.
In 2016, the bottom fell out. A county that had given 52 percent of its vote to Barack Obama in 2008 gave just 41 percent of its vote to Hillary Clinton—changing the vote margin from +14,000 Obama to +34,000 Trump. It only got worse in 2020 and 2024. This year, Trump won it by nearly 68,000 votes.
In 2018 and 2020, we saw our weakness with African American and Hispanic voters. Black turnout was down both cycles, and in 2020, Hispanic support was down significantly.
At Unite the Country, we saw some of these issues among Hispanics even among the most loyal Democrats in primary states. Polling in both Illinois and Florida showed Democratic Hispanic primary voters starting to react negatively to the socialism talk that came out of our 2020 presidential primaries—so much so that we did pro-Biden mail calling out the “revolution” posture coming from some of the presidential candidates in both states to make it clear he was focused on bread-and-butter issues.
Still, despite plenty of people screaming from the rafters, the Biden campaign largely ignored the growing problems associated with the rhetoric of the extreme left. In doing so, it let the narrative settle in. Trump’s gains among Hispanics in Florida were significant: from +27 Clinton to +7 Biden among Hispanics. Because the shift among that electorate was just three points nationally, national Democrats largely wrote it off as an outlier—Florida being Florida, and all.
But eventually, we saw that Florida hadn’t been an outlier at all.
“Smart” Campaigns Are Making Us Dumb
HAS ANYONE HAD TO BUY A refrigerator recently? After having previous had one for twelve years, I’m now on my third in seven—not because the actual refrigerator is breaking but because all the “smart” technology on my fridge is. Last time I went in, I said, I just want a dumb fridge that works. Just the basics—works for appliances and for elections.
I came to politics by accident. My first boss, Doug Wiles, represented one of the most Republican state legislative districts held by a Democrat in Florida, so I learned the key to winning, besides turning out your base, was to keep the margins as close as possible in the areas you are going to lose.
Politics isn’t rocket science. It is just fourth-grade math. You take the likely voter turnout, divide it by two, and add one. Get to that number, you win. It really is that simple.
The nature of the Democratic coalition means most victories in a battleground state or district allow you to win fewer places than you lose. How we manage margins, in turn, matters.
I liked to tell our field organizers working in red areas in ‘08 and ‘12 they were just as responsible for the floor as they were the ceiling. If the floor in red precincts gets too low, the math to make it up in blue ones often doesn’t exist.
Take Florida in 2016. Hillary Clinton won the base Democratic counties by more than Barack Obama, yet lost. Why? She lost the red ones by substantially bigger margins than Obama. The new floor was so low it didn’t matter that she had set a new ceiling in our base communities.
The collapse of our floor didn’t just happen in Florida. Obama carried Polk County, Iowa—home to Des Moines—by 14 points in 2012. In 2024, Harris won it by 11. But compare that to Woodbury County, home to Sioux City. Obama won it by less than 1 point in 2012; Harris lost it by 23. It’s hard to win states when huge numbers of counties move 20 or more points to the right.
Add moves like this to the shifts we saw with African American and Hispanic voters, and the long-term winning coalition is pretty much gone.
Truth be told, thanks to “smart” election technology—in this case, campaign analytics and modeling—we increasingly don’t talk to voters in large swaths of states.
I am not anti-data. All data are useful. The problem with using analytics to “efficiently” target voters is that efficiency leads to mass exclusion.
For example, in 2008, in many southern states, the Obama campaign sent mail on gun policy to men in rural and exurban counties. We laid out his positions: that he believed in background checks, and closing the gun show loophole. We didn’t expect to win those voters on guns but we knew we needed to take our position directly to them in order to get them to listen to us on other issues.
There is zero doubt in my mind that current modeling would tell us to never do this, that it would be crazy inefficient.
We’re over-reliant on analytics and not on the emotional or narrative elements of politics. A lot has been made of the Trump campaign’s anti-trans ad attacking Harris for defending gender-affirming care for federal prisoners. That ad wasn’t about elevating trans issues. It was about making Harris’s priorities look crazy and out of touch.
The Harris campaign decided not respond, in part because the ad testing said it wasn’t compelling, and because they didn’t want to play on their own side of the field. You know what happens when you don’t play defense on your own side of the field? I am a Jaguars fan, just watch some of our games.
The problem with not responding is that the issue settles in and, for many voters, it becomes perceived as truth. Surveys conducted after the 2024 election showed that voters thought trans issues were the second-most important issue to Democrats.
Sometimes communicating by gut and not data is essential. As a senior staffer from a campaign where a Democrat had won a governorship in a red state told me: “You know where we didn’t test a single ad? In the race we won.”
The Median Voter Will Lead the Way
I KNOW THIS HAS BEEN A LONG piece, but there is so much more I could say.
And there are a lot of obvious things I didn’t say. Yes, we need to do better with men, we need to engage more younger voters, we need to reverse the absolute collapse we’ve seen with Hispanics, and we need to get back to persuading black voters, not taking them for granted.
But basically, we must get back to listening to the median voter. Bill Clinton got this. Barack Obama got this. Donald Trump gets this. That median voter probably isn’t the white “soccer mom” or “security mom” or “NASCAR dad” of elections past. Rather, the median voter today is just as likely to be a 38-year old African American father trying to figure out how to raise two kids, or a 28-year-old Hispanic woman struggling with affordable housing.
I find a lot of my party’s handwringing over where policy should fall on the ideological spectrum to be misplaced, because all that really matters is if we can convince that median voter. Win 50 percent of those voters plus one, and you win most elections.
Increasingly, our side believes we should only communicate on issues where voters give us an edge. But when the median voter isn’t there, or is worried about other issues, our communication echoes in a void. It is no wonder so many voters wonder what the hell we are all about.
Going forward, the map is going to change. The “Blue Wall” states in the industrial Midwest will likely not be enough to reach 270 Electoral Votes after the 2030 census. To build any kind of sustainable majority to elect a president, our coalition must evolve. We must win states we traditionally lose, meaning we must do a better job of listening to, and eventually persuading voters whose world views are different than the coastal leaders of our party.
We’ve seen what happens when we don’t listen to voters—when we focus on reinforcing our tent instead of expanding it, and when we move our message outside of the mainstream.
And in my home state of Florida, we’ve seen what happens when we stop listening. Without course corrections, more and more states are going to live through what I have lived through for the last eight years.
I have no doubt Trump will over-read his narrow mandate and overreach. It’s a tradition among most presidents. This is an opportunity for my side to redefine our values for voters who have stopped listening. Get this right, and we set ourselves up nicely for the next decade. Get this wrong, and we could be in the wilderness for a very long time.