Democrats could be forgiven for dreaming about a "blue" Florida. It is diversifying as fast as Texas or Arizona, and the demographic composition of its electorate may be poised to shift even faster than anticipated.
As many as 300,000 people have fled to Florida from Puerto Rico in the aftermath of Hurricane Maria. And a ballot initiative this November could return the vote to the state’s estimated 1.5 million discharged felons. At first glance, either tally of these two Democratic-leaning groups would seem to dwarf Donald Trump’s 113,000-vote margin of victory in the state in 2016.
But the reality for Democrats is that neither development is likely to fundamentally alter Florida’s political character heading into the 2020 election.
The main reason? The electoral effect dwindles after accounting for the relatively low turnout rates among these groups. More generally, even big demographic shifts that seem to favor Democrats could easily be swamped by other demographic shifts that do the opposite.
Of the two major shifts — assuming the ballot measure clears the 60 percent approval required, and that’s hardly a sure thing — the influx of Puerto Ricans is probably less likely to have a major effect. That’s even if you accept the high-end estimate of 300,000 new residents (and some university professors have questioned that total).
To start, exclude the roughly 75,000 children out of that 300,000 figure, based on the 24 percent of Puerto Ricans who were younger than 18 in the 2010 census. Then consider how many eligible adults will actually register to vote. Let’s generously assume that 57 percent will register by 2020, the same share for Hispanic adult citizens nationwide, according to the 2016 census Current Population Survey. That would mean around 130,000 newly registered voters, not all of whom will vote.
According to Daniel Smith, a professor of political science at the University of Florida, about 62 percent of Florida’s registered Puerto Ricans voted in 2016, a lower percentage than that of other Hispanic groups in Florida. That would mean around 80,000 votes, and not all of those voters will support Democrats. Even if Democrats won them by a big margin, 75 percent to 25 percent, for example, they would still net only around 40,000 votes.
Holding everything else constant, a net gain of 40,000 voters would have cut Trump’s margin of victory to 0.8 points from 1.2 points. That could well be crucial in a close election — ask Al Gore — but a 0.4-point shift would not require a real re-evaluation of Florida’s politics, or come close to meriting a declaration that it was shifting "blue."
Certainly, no one would be discussing Florida vastly differently if Trump had won it by 0.8 points instead of 1.2 points.
The re-enfranchisement of most former felons in the state is potentially much more significant. And while there’s considerable evidence that ex-convicts have a low turnout rate, they don’t need to have a very high turnout rate to make a difference if there’s roughly 1.5 million of them.
Florida is one of a few states with a permanent voting ban for people with felony convictions, a policy that goes back to the Jim Crow era. A discharged felon can apply individually for clemency to the governor, but Florida’s rules are restrictive. The ballot initiative would automatically allow felons to vote once they completed their sentences, except those convicted of murder or a felony sexual offense.
Ex-convicts tend to have very low educational attainment, and they tend to be young and nonwhite, with low incomes — all factors that correlate with a low turnout. The ex-convict turnout rate might not be expected to be much higher than 30 percent, based strictly on demographic characteristics.
There’s evidence the actual turnout rate among this group is even lower. Over the last decade, several studies have matched discharged felons to voter registration files in the states where they’re permitted to vote. Most studies find that around 20 percent of them turn out in battleground states in presidential elections.
Many of these studies were conducted with data from the 2008 and 2012 presidential elections, when black and youth turnout were particularly high and Democratic registration rates were particularly high as well. On the other hand, these studies were of recently discharged felons, who might be less likely to register and vote than the broader felon population.
A 20 percent turnout rate is pretty low, but it could produce a meaningful effect over 1.5 million newly eligible voters. It would mean 300,000 new voters — more than three times the estimated number of new Puerto Rican voters.
The same studies indicate that felons were highly likely to register as Democrats (African-Americans, a strong voting bloc for Democrats, are a disproportionate share of the disenfranchised.) That would probably be enough to cover Trump’s 113,0000-vote margin of victory from 2016, and perhaps quite a bit more.
But it probably wouldn’t mean a significant and durable edge for Democrats. Just ask Hillary Clinton, who inherited a 90,000-vote advantage from Barack Obama in the 2012 Florida election — President Barack Obama defeated Mitt Romney, 50 percent to 49.1 percent — and seemed to benefit from four years of demographic shifts, including an influx of Puerto Rican voters, and still lost the state by more than 100,000 votes.
One reason is the overlooked influx of fairly conservative white voters. Florida has long been a retirement destination, and its above-average number of high-turnout older white voters have both trended toward the Republicans and helped counter other demographic shifts.
The Villages area near Leesburg was the fastest-growing metropolitan area in the United States in 2016; it voted for Trump by 39 points. Overall, 10 of the 25 fastest-growing metropolitan areas in the United States were in Florida in 2016, and all except Orlando-Kissimmee — where there has been considerable Puerto Rican immigration — supported Trump in 2016.
In 2017, newly registered Floridians registered as Republicans by a margin of 29 percent to 27 percent over Democrats, according to data from L2, a nonpartisan election data vendor. That’s despite the uptick in Hispanic registration after Hurricane Maria hit Puerto Rico on Sept. 20 (Hispanic voters represented 22 percent of registrants in the two months before Maria, and 29 percent in the two months after). White voters over age 50 registered as Republicans by a margin of 45 percent to 21 percent.
Another reason: Seemingly big demographic changes can easily be swamped by even slight shifts among the rest of the electorate. If Trump were to gain another percentage point among white working-class voters, for instance, that could be enough to overcome the combined effect of both new Puerto Rican residents and felon re-enfranchisement. Persuading a voter to switch sides is twice as helpful as turning out a new one: Persuasion both takes away a voter from your opponent and gives you one. Adding a new voter does only the latter.
Over the last two decades, Florida has been the closest state in the country in presidential elections. That may change one day. But if it does, it probably won’t be because of a single, modest demographic shift, like felon re-enfranchisement or an influx of Puerto Rican voters. Florida is the nation’s third-most populous state. It takes a lot to move the needle.
Nate Cohn covers elections, polling and demographics.
© 2018 New York Times