Want the Real “New” Voter Registration Numbers in Florida?

Matt Dixon of the Florida Times Union, who I think is a fine journalist, had a piece the other day with the headline: Democratic registration all but dries up since new Florida laws.

The column has been picked up several by several outlets, including the New York Times and Rachel Maddow’s Blog, and it has sparked lively discussions in such outlets as Weasel Zippers, Hullabaloo, and Addicting Info.

But Dixon’s figures, which I discussed earlier today here, are wildly off the mark and completely misleading.

Yes, “new” voter registrations are down in Florida, but nowhere near are they as dire as Dixon’s column suggests.

My collaborator Michael Herron at Dartmouth and I have just crunched the numbers, using data from the Florida Department of State voter files created on April 1, 2012 and April 1, 2008.

According to the state’s official records, more than 155k voters registered as Democrats in 2011.  Slightly less than 138k voters registered as Republicans in 2011.

Not surprisingly, given House Bill 1355’s draconian restrictions placed on third party voter registration organizations (3PVROs), which went into effect on July 1, 2011 and frustrated the ability of groups like the League of Women Voters to sign up new voters, registration figures for both parties were down in 2011 compared to 2007 (28.1% fewer Floridians registered as Democrats in 2011 compared to 2007, and 15.5% fewer Floridians registered as Republicans in 2011 compared to 2007).

But these dampened voter registration figures are nowhere near those published in Dixon’s column.

The bottom line: More than 155 thousand voters registered as Democrats and more than 138 thousand voters registered as Republicans in Florida in 2011.  The numbers are down, but that hardly looks like “drying up” to me.

It’s Time to Get the Voter Registration Numbers Correct in Florida

Florida’s controversial election law, HB 1355, which has restricted early voting, made the casting of provisional ballots more likely, and cracked down on third party voter registration organization (3PVRO) efforts to register eligible Florida citizens, is still making news.

Yesterday, Judge Robert Hinkle, a federal judge in Tallahassee who is presiding over the legal action brought forth by the League of Women Voters and the other plaintiffs challenging the constraints and penalties placed on 3PVROs, took a more definitive step in deep-sixing several provisions of Governor Scott’s signature voter-suppression law, ruling that he intends to issue a permanent injunction as soon as the 11th U.S. Circuit Court of Appeals dismisses the state’s appeal of his preliminary injunction that he issued back on May 31, 2012.

Anecdotal evidence from groups sitting on the sidelines in Florida suggests that the law has indeed dampened voter registration across the state.

But how should we measure such a decrease, if indeed there has been one?

Florida Times-Union reporter, Matt Dixon, tried to measure the impact of the law in a story he wrote earlier this week that has gotten a lot of press, including a reprint of his findings in the New York Times and on Rachel Maddow’s Blog.  The headline of his article screams, “Democratic registration all but dries up since new Florida laws.”

But let’s take a closer look at Dixon’s methodology.

According to Dixon’s analysis, between July 1, 2011 and August 1, 2012, the number of registered Democrats statewide increased by only 11,365, compared to increases in Democratic registrations over the 13 months that preceded the 2004 and 2008 elections, which he claims increased an average of 209,425 voters.

Dixon also provides similar numbers of new registrations for Republicans, as show in this chart produced on Maddow’s Blog:

Chart: Florida’s voter-registration collapse

Although voter registration numbers are indeed down in Florida compared with other years, there are serious problems with Dixon’s analysis.

First, it does not measure NEW voters who are registered over the three 13-month periods.  Rather, his analysis looks at the differences in aggregate registration numbers for the two parties, which does not hold constant any increases, or (in fact) decreases in the overall number of registered voters in Florida.  Furthermore, statewide population had increased prior to the two previous elections, but recently has remained flat, decreasing the overall pool of potential new registrants of either party.

Second, voter registration is a daily, ongoing occurrence, and total number of registered Democrats and Republicans Dixon compares over the three spans does capture the possible impact of HB 1355, which went into effect on July 1, 2011. Indeed, his 13-month time-frame includes two months (June & July 2012) when 3PRVROs in Florida were once again registering voters after Judge Hinkle’s preliminary injunction on May 31, 2012.

A much better way to measure the effects of HB 1355 on voter registration numbers in Florida is to do what Professor Michael Herron and I have done in this paper.

I don’t have time to summarize our results, but perhaps these figures might suffice for now:
As we write on page 19 in our paper,
I’ll certainly have more on this later…

Exclusive Empirical Research: “House Bill 1355 and Voter Registration in Florida”

Professor Michael Herron (Dartmouth College) and I have posted a draft of our American Political Science Association annual conference paper, “House Bill 1355 and Voter Registration in Florida,” here.

Here’s the Abstract:

New state laws governing voter registration went into effect in Florida on July 1, 2011. Among the legal changes
promulgated as a consequence of a piece of Florida state legislation known as House Bill 1355 were new registration
requirements for third-party groups like the League of Women Voters and a new oath, warning of prison time and fines,
that voter registration agents had to sign before engaging in registration activities. Such changes raised the implicit
costs that eligible Florida citizens faced when registering to vote, and we show, consistent with this logic, that voter
registrations across Florida in late 2011 dropped precipitously compared to registrations in late 2007. This pattern is
evident among registrants in general, among registrants age 21 and younger, and among the number of individuals
who registered as Democrats as well as the number who registered as Republicans. Outside of House Bill 1355, we
know of no credible explanations for our findings about Florida registration drops in 2011. Our results thus show how
restrictions on the way that third-party organizations register voters can have tangible effects on actual registrations
and, given that registration prior to an election is a civic necessity in Florida, can affect electoral participation.

EXCLUSIVE: The Effects of Florida’s HB1355 Early Voting Law on Turnout

As was reported widely in the press, if not entirely accurately, last Thursday night a Washington, DC, panel of federal judges handed down a unanimous ruling that restrictions placed on early voting in Florida should continue not to be implemented in the five counties covered by Section 5 of the 1975 amendments to the 1965 Voting Rights Act.  Florida, said the panel, “has failed to satisfy its burden of proving that those changes will not have a retrogressive effect on minority voters.”

With respect to early voting, House Bill 1355–which was passed on party line votes by the Florida legislature and signed into law by Republican Governor Rick Scott in May 2011–is likely to have a differential impact on racial and ethnic minority voters in the 2012 general election.  In addition to my testimony before the US Senate on the topic, I’ve co-authored with Professor Michael Herron of Dartmouth College a soon-to-be published article in Election Law Journal that reveals the heavy reliance of early voting by minorities in the 2008 general election.  We found that in the 10 Florida counties that offered voting on the final Sunday of early voting in 2008, there was a surge in turnout among minority voters, especially African Americans.  That final Sunday of voting was eliminated under HB1355.

Since then, Florida voters have participated in two statewide primary elections in 2012 under a dual system of elections, which very well may violate state law.

Because the Secretary of State decided to enforce HB 1355 in 62 of the state’s 67 counties, despite the fact that the US Justice Department refused to preclear the enforcement of HB 1355 for Florida’s  five covered counties (Collier, Hardee, Hendry, Hillsborough, and Monroe) under Section 5 of the VRA, those five continued to offer two weeks of early voting, (Monday through Saturday, and again, Monday through Saturday) for a total of 96 hours. (Incidentally, both prior to and after the enforcement of HB 1355, the elections supervisors of each of the five counties opted not to offer voting on either of the two Sundays, instead allotting the required eight-hours of weekend early voting all on the two Saturdays).

In contrast, the state’s other 62 counties were required by HB 1355 to cut back on the total number of days of early voting (from a maximum of 14 days, Monday through Sunday, and again, Monday through the final Sunday before the election; to a maximum of eight days, Saturday through Saturday).  Under the new early voting restrictions contained in HB 1355, which apparently eluded both Governor Scott as well the Chairman of the Republican Party of Florida, Supervisors of Elections and could offer as few as 48 total hours (but no more than 96 hours over the truncated period).

As I noted back in March following Florida’s (Republican) Presidential Preference Primary (PPP), under Florida’s dual election system the percentage of early voters casting a ballot averaged across the five Section 5 counties was higher than the average use of early voting in the state’s other 62 counties, with the reduced days and hours of early voting.  As I wrote then, “a greater percentage of registered Republicans opted to vote early in-person in the five Section 5 counties than registered Republicans in neighboring counties. Some 11.8% of registered Republicans voted early in the five VRA Section 5 counties, compared to 9.3% of registered Republicans who voted in the other 62 counties.  More significantly, early in-person voting in the five counties with the extended voting window accounted for a greater percentage of the total turnout in the Presidential Preference Primary, on average, compared with turnout in the other 62 counties.  Nearly one in three votes cast in the GOP primary election in the five Section 5 counties were ballots cast early in-person by voters, compared to less than 22% of all ballots cast in the other 62 counties. Florida 2012 Presidential Preference Primary Section 5 VRA Counties Early Voters Graph

The same was true in the recent August 14 statewide primary with Republicans, Democrats, third party, and nonpartisan voters coming to the polls, as Dr. Herron and I found by merging the 67 county early voter files (which we downloaded three days after election day) with the state’s (yet to be official) turnout figures from the 67 counties.  Of the roughly 183k registered voters who turned out to vote in the five Section 5 counties (Collier, Hardee, Hendry, Hillsborough, and Monroe), 18.96% came to the polls in person over the two-week early voting period (Monday-Saturday, Monday-Saturday). (As I mentioned above, each of the Section 5 counties still operate under the old law, and in addition to offering eight hours a day Monday through Friday, have opted to offer eight hours on the two Saturdays, for a total of 96 hours of early voting).

In contrast, of the nearly 2.16 million registered voters who turned out to vote in the state’s other 62 counties, which had a total of only eight days of early voting (Saturday through Saturday) and averaged considerably less than 96 total hours over the week, only 15.39% voted early, nearly 3.6 percentage points less than the five Section 5 counties.

This certainly seems to provide further evidence that the DC federal court last week was correct in striking down Governor Scott’s effort to curtail early voting.  Having a longer time-frame, with more available hours to cast an early ballot, only enhances a citizen’s likelihood of getting to the polls–including African Americans living in the protected counties (although we need to conduct further analysis to verify this possibility).  But of course, that’s why most Republican lawmakers supported HB1355–and Governor Scott quickly signed it into law–because they’re not terribly interested in trying to expand Florida’s electorate.

And early voting may very expand the electorate, notwithstanding the conventional wisdom (expressed even by those who opposed the cut-back of early voting, including Reed College Professor Paul Gronke, who served as an expert opposing the shortening of early voting in the DC litigation) that early voting may simple redistribute the timing of those citizens who already plan to cast a ballot vote.

Florida’s quasi-natural experiment, which Dr. Herron and I are investigating for a series of papers (including one we’re presenting next week at what promises to be a fantastic panel on the strategic mobilization of minorities at APSA, with Dr. Matt Barreto serving as our discussant), provides some leverage into the question of whether HB 1355 may have a depressing effect on overall turnout.

What we’re seeing in Florida (due to the state’s dual election system), is that having early voting spread over two weeks appears to give a wider range of voters more opportunities to turn out to vote.  When faced with fewer days and shortened hours of early voting, registered voters who have become habituated to vote early may decide to stay home.

This is particularly an important point when thinking about the five Florida counties covered by Section 5 of the VRA.  Average turnout is historically lower in these five counties than the statewide average. This, of course, is one of the reasons why the five counties were added in 1975 to those counties protected by the Justice Department under Section 5 VRA.

Average turnout of the five Section 5 counties compared to the state’s other 62 counties over the five statewide elections immediately prior to the implementation of HB 1355 (2008 PPP, 2008 primary, 2008 general election, 2010 primary, and 2010 general election) was 4.61 percentage points less, 40.81% to 45.42%.

And average turnout of the five Section 5 counties over the two statewide primary elections held in 2008 (the January PPP and the August primary) was only 30.13%, compared to an average of 35.79% in the state’s other 62 counties, an even greater difference of 5.66 percentage points.

But when comparing turnout numbers for the two primaries (the January PPP and the August primary) held in 2012 under Florida’s dual election system–with more early voting opportunities for registered voters living in the five section 5 counties compared to those residing in the other 62 counties–the turnout gap narrowed considerably, on average.  Turnout in the five Section 5 counties was 32.97%, compared to 36.20%, a difference of only 3.23 percentage points.

So, what are we to make of this? Admittedly, the turnout gap between the five Section 5 and the other 62 counties is not huge, but it is indicative that HB 1355 may be depressing turnout in those counties that must comply with the new, more restrictive law.  And, it is certainly arguable that since registered voters in the five Section 5 counties have historically relied more heavily on early voting in past elections, and if early voting days and hours are reduced in those counties if HB 1355 is eventually upheld, their comparatively lower turnout levels might take even more of a hit.

Unfortunately, we will not have a more definitive answer to this possibility until after the 2012 election.  But with what we’ve seen so far, it seems pretty clear that reducing the number of days and hours of early voting in Florida has had a negative effect on voter turnout.

Stay tuned. Much more from Dr. Herron and me on this later…

Still Confused about Florida’s Early Voting Law? You’re Not the Only One

I just took a quick stroll around the web to check out the headlines announcing the DC federal district court’s decision yesterday on Florida’s 2011 early voting law.

Every one of the following headlines, except one, is misleading.


Federal court rejects Florida early voting changes

Rick Scott Strikes Out Again: Federal Court Blocks Florida Attack On Early Voting

Federal court says early voting cutback could hurt blacks

Federal Court Reinstates Early Voting Days In Parts Of Florida

Florida judge dumps early voting limits in some counties

Black Vote: Federal Court Reinstates 4 Early Voting Days in 5 Florida Counties


Not sure which one? Check out yesterday’s post where I try to explain Florida’s early voting law, before and after the passage of HB 1355.