OK, there’s been a tremendous amount written on voter registration trends, by party, in Florida. It’s all wrong. Here’s why.

I’ve published quite a bit on the topic, served as an expert witness in dozens of successful lawsuits, and know these files inside and out.

Here are the facts, using proper voter files (i.e., contemporaneous), comparing Jan-July 2016 registration numbers with Jan-July 2020 voter registration numbers, comparing apples to apples.

To look at new registrations (that is, how the parties and groups are doing on the ground), you cannot merely use the Florida Division of Elections summary files to figure this out, as I explained four years ago, here.

OK, here are the numbers. Comparable voter registration, as of July 1, 2020, was down by a total of 136,392 registrants. And Democrats and Republicans both account for 32.1% of new registrants over the first six months of the year.

New Registrations by Party, January-June, 2016 vs. January-June 2020

Dem Rep NPA 3rd Total
Jan-June 2016 152,320 131,059 149,090 10,273 442,742
34.4% 29.6% 33.7% 2.3% 100.0%
Jan-June 2020 98,439 98,472 99,879 9,560 306,350
32.1% 32.1% 32.6% 3.1% 100.0%
Difference (2020-2016) -53,881 -32,587 -49,211 -713 -136,392
-2.3% 2.5% -1.1% 0.8%

 

So, cut these numbers as you will, but by no means can it be said that Democrats are killing it out there.  In fact, a total of 33 more Republicans registered to vote anew between January 1 and June 30 (and remained registered as of June 30) than newly registered Democrats.

COVID-19 is obviously the major reason why Democratic numbers are down, as nonpartisan groups who usually hit the ground are not as active, SOEs who usually are in the schools preregistering young voters had no classrooms to go to in April or May, etc.  We’ve seen this pattern before, in 2011, after the Republican legislature passed HB 1355, which, among other things, cracked down on voter registration efforts by 3PVROs, as Michael Herron and I have written about before.

Perhaps Democrats and their allies will turn it around before the October registration date. Young voters who register immediately before an election in Florida are more likely to turn out in that proximate election, as Enrijeta Shino and I find, but their turnout levels aren’t sustained in subsequent elections, casting some doubt on the habitualization of voting.

But until we get some book closing numbers, the data are what they are. Republicans, as a proportion of new registrants in Florida, are doing better than they did four years ago over the first six months of the presidential election year.

Young people are marching after #Parkland, but are they registering at a higher clip in Florida?

Let’s take a deeper dive into new registered voters in Broward County, Florida. According to a story in the New York Times by & , young people are registering. It’s hard to tell from their graphic (below) whether the total new registered voters in Broward in the month of March is 3,416 (628 + 626 + 1575 + 587) or if the figures are for just 18-25 yr olds newly registering to vote in the county.

NYT Broward Young Registration

According to my count, which draws on the May 2018 statewide voter file, some 4,383 voters registered anew in Broward County in March 2018. Of those, 2,026 (46%) are 18-25 years old.

But what do these new registration figures mean?

Here are the comparable March 2014 new registrations in Broward County (that I’ve drawn from a June 2014 statewide voter file).  In March 2014 in Broward, there were 9,853 new registrations, more than twice as many my figures indicate were recorded in March 2018.  Of those in March 2014, 2,972 (30%) were 18-25 yrs old (as of June 6, 2014).

So, compared to four years earlier, a higher percentage of new registrants in Broward in March 2018 are aged 18-25, but that’s because half as many people registered in March 2018 compared to March 2014 in the county.  In March 2014, nearly 1,000 more 18-25 year olds than in March 2018.

Of course, there’s always some slippage when dealing with voter files. My snapshot for 2014 is from early June; it’s certainly possible that some individuals who registered in March 2104 and who were living in one of the state’s other 66 counties moved to Broward in April or May, upping Broward’s June count. It also appears that the March 2014 count of new registrations in the statewide voter file includes thousands of “pre-registrants,” that is, individuals who were 16 or 17 years old who registered to vote that March. Those individuals don’t appear in the March 2018 statewide voter file.

What about the partisan breakdown of younger registrants across the four year time span?

Surprising no one, in Broward County this past March 2018, of the more than 1,800 newly registered voters 18-21 years old, 48% registered as Democrats, 37% as NPAs, and 12% as Republicans.

Statewide in Florida, in March 2018 some 10k 18-21 year olds newly registered to vote. Roughly 38% are Ds, 19% are Rs, and 40% are NPAs.

For a comparison of new party registrations, statewide in March 2018 there were slightly less than 11k new voters aged 45-64. Of these older, new registrants, 26% Ds, 33% Rs, and 38% NPAs.

None of this is news. Younger voters register to vote. They are more likely to register as Democrats. Older people newly register, too. In substantial numbers. They are more likely to register as Republicans.

So, a couple more comparisons of new registrations in Florida in 2018 versus 2014.

Between January 1 and April 30, 2018, there were a total of roughly 155k new registrations statewide. Of those new registrants, 50.2k are 18-29 year-olds and 44.6k are 45-64 year-olds. Of the 18-29 group, 32% are Ds, 20% are Rs, and 45% are NPAs. Of the 45-64 group, 26% are Ds, 34% are Rs, and 37% NPAs.

Compare these figures with first four months of 2014. Between January 1 and April 30, 2014, there were 166.6k new registrations in the state. Of those, 50.2k were 18-29 years old.  Another 22.4k were 45-64 years old. Of the 18-29 group, 28% were Ds, 19% were Rs, and 48% NPAs. Of the 45-64 group, 29% were Ds, 33% were Rs, and 33% were NPA.

One last point about the supposed causal relationship between the uptick in voter registration among younger individuals and Florida and the #Parkland tragedy.  The Times reporters write:

In Florida, voters under 26 jumped from less than 20 percent of new registrants in January and February to nearly 30 percent by March, the month of the gun control rallies. That ticked down to about 25 percent in April, as the demonstrations subsided, but registration of young voters remained above the pace set before 17 students and faculty were killed at Marjory Stoneman Douglas High School in Parkland.

It’s amazing that the exact same pattern existed four years earlier, over the first four months of 2014, when Florida did not have a school massacre and there weren’t mass demonstrations against gun violence and the #NRA.

The percent of total new registrants aged 18-24 statewide in Florida in January and February 2014 was 20.6% and 20.3%, respectively. In March, 2014, the percentage of all voters registering in the state aged 18-24 jumped to 34.5% — even higher than the percentage in 2018 following #Parkland and student mobilization.  Then in April, 2014, the 18-24 group’s percentage of new registrants dropped down to 24.6% of all new registrants, which was “above the pace set before…….”  You get the picture.

This is really a story of omitted variable bias. There are lots of other reasons why young people register in March — starting with ramped-up voter registration drives by the state’s 67 Supervisors of Elections in High Schools and at community events. Many of the students and other young people in Florida who were understandably and genuinely mobilized to register after the Parkland shootings would have registered anyway at one of these events, just as many of their older brothers and sisters did four years earlier.

In short, it’s still way too early to draw conclusions about the supposed registration effect of #Parkland.

Of course, turnout of younger voters in a midterm election is another issue altogether, but that story will have to wait for another day.

Think those last-minute registrants who vote in the upcoming election will turn out in future elections? Think again.

My latest research, coauthored with UF PhD candidate, Enrijeta Shino.

Timing the Habit

Online version available here.

 

Will Influx of Puerto Ricans flocking to Florida hurt Trump or Republicans?

I merged information about the birthplace of the 13m registered voters in Florida with precinct results in the 2016 General Election.  The following figure reveals the two party vote for Trump and Clinton in those Florida precincts that had at least 100 Puerto Rican-born registered voters who voted in the 2016 General Election. Each of the precincts’ voters were at least 50% Hispanic, and at least 1/5 of those Hispanics who voted were Puerto Rican-born.  In other words, these are about as Puerto Rican of precincts in Florida as it gets.

For example, in these majority-Hispanic voter precincts, in which, say (on the X-Axis), 30% of Hispanics who voted were Puerto Rican-born (and that have at least 100 PR-born voters), Trump won only 30% of the vote, on average. He did very poorly in all of the predominantly Puerto Rican precincts.

PR 2016 Vote

Does this mean that Republicans are in trouble in FL if the exodus from PR to the mainland happens? Puerto Ricans, certainly compared with other Hispanic groups in Florida, have weak voter turnout.  Of the more than 180k PR-born voters in my database, only 112k of them voted in 2016 (62%). That’s much lower turnout rate than, say, Cuban-born voters in FL. In 2016, 242k of the 325k Cuban-born naturalized citizens in my database turned out; that’s roughly 75% turnout.

A Manic Morning ElectionSmith Exclusive: Party, Race/Ethnicity, and Age Breakdown of >90k Floridians Who Successfully Registered to Vote During Extended Week following FDP Lawsuit

With an extra hour from daylight savings, I was able to crunch some new numbers…

According to the Florida Department of State, Division of Elections transparent by rather cryptic website, over 107k voters registered between October 11 and October 28.  According to the DSDE’s website, “Pursuant to a court order, Florida’s voter registration deadline for the 2016 General Election was extended from October 11 to 5:00 PM on October 18. The Department of State’s Division of Elections is committed to ensuring that all eligible Floridians who submitted a complete voter registration application by the extended voter registration deadline can vote.”

As the expert who crunched the numbers for the plaintiffs in that federal lawsuit (Florida Democratic Party v. Richard Scott) I’m obviously interested in the numbers of new registrants.

For some reason, DSDE’s website of daily activity reports provides only the “total number of voter registration applications verified and made active, and voter registration applications in the verification process, on or after October 11.”

Here’s the final running count provided by DSDE.

Date Applications Verified & Active (Registered Voters)
10/11/16 – 10/28/16 (4:12 PM) 107,085

I’m not sure why October 11 registrations are provided as the start date of the running total, as that was the regular final day (bookclosing) to register to vote in Florida prior to the court’s extended week of VR.  The DSDE’s processed voter registration dates also run through October 28; these voters who were processed after the October 18 extension should be eligible to cast regular ballots in the November 2016 General Election, as the DSDE states that, “The Department will be providing updated daily numbers until the state provides statistical bookclosing (voter registration) reports for the upcoming General Election.” [emphasis added]

Below are the party and demographic breakdowns of the 90,695 Floridians who were successfully processed by the state’s 67 county SOEs and the DSDE and recorded in the statewide voter file as being registered to vote on October 12 (the first extended day per the court’s order) through October 18, 2016.

PARTY

Democrats: 28,357 (31%)

Republicans: 23,410 (26%)

NPAs: 36,414 (40%)

3rd Parties: 2,514 (3%)

 

RACE/ETHNICITY

Black: 10,782 (12%)

Hispanic: 19,239 (21%)

White: 45,050 (50%)

All Others: 15,624 (17%)

 

AGE

18-29: 32,448 (36%)

30-44: 22,939 (25%)

45-59: 18,042 (20%)

60-105: 13,810 (4%)

Other (Pre-Registrant; Over-105; Redacted): 3,456

 

These October 2016 figures comport to my expert report in the FDP v. Scott lawsuit, where I find that in 2012, the more than 116,000 individuals who successfully registered to vote in the final six days before the October 9, 2012 bookclosing were disproportionately more likely to be racial/ethnic minorities and younger (I was not asked to look at the party of the registrants by the plaintiffs).

Gaining a hour of sleep (work) feels pretty good, but not nearly as good as I feel having done my part to extend the voter registration deadline in Florida.  Pretty amazing what a little tweet can do…

scott_vr_matthew