Alabama Continues to Marginalize African-American Voters

In 2010, Alabama’s state legislature fell into Republican hands for the first time since reconstruction.  The long-held Democratic majority was replaced with a Republican super-majority in one night.

Alabama House Alabama Senate

The Republicans of Alabama wasted little time working to cement their hold on power.  They gerrymandered the legislative boundaries to pack in African-Americans, a move so blatant the Supreme Court of the United States ordered the lower courts to review the maps on a case by case basis.  In the redistricting maps, 1/6 of all African-Americans were moved from white districts to black districts, in many cases making black districts over 70% non-white.The goal was simple, take African-American voters out of swing seats and pack them into minority-access seats.  The issue is going through the lower courts again.

The legislature also passed another measure to cement their new majority.  They passed a voter ID law, requiring any Alabama voter have a license to cast a ballot.  These laws sound good but are shown to disproportionately effect the poor.  They also show no evidence of stopping voter fraud, which is shown to be a very rare occurrence to begin with.  In fact, many instances of voter fraud involve absentee ballots, which ID laws can’t prevent.

Nevertheless, the voter ID law passed in Alabama.  Then just recently, the state announced that due to budget cutbacks,it was closing 31 drivers license facilities, leaving 29 counties without a venue to acquire a drivers license.  This is the main form of ID for a large majority of Alabama voters.

It should be noted that many of the rural and urban poor of Alabama do not have the resources to drive into another county for a license, or the internet access to purchase one online.

What has caused this story to gain national attention is the fact that the facility closure’s disproportionately effect African-American counties.  The map below shows the African-American share of the counties and which counties will be left with no license facility.


Many of the counties without a facility reside in the black-belt, the largely African-American, and reliably-Democratic, region of the state.  With these closures, only 3 of the 11 counties over 50% African-American will have a facility order licenses.  My graph below sorts the counties of Alabama by African-American % and color-codes which county will have or not have a license facility.  The blue line is Obama’s 2012 share of the vote.

AlabamaDMV Chart

The top six African-American counties will have no facility at all.  13 of the 29 counties with no facility are 40% or more African-American.

The state of Alabama’s plan disproportionately effects African-American counties, most of which have higher poverty-levels and cannot easily travel to another county.


Poverty statistics and a high resolution version of the above map/charts can be found here.

A lawsuit will no doubt follow this decision and it is hard to see a scenario where the justice department doesn’t get involved.  Only 26% of Alabama is black but 8 of the 11 majority-black counties will no longer have a facility that helps the citizens comply with a voter law passed just a few years ago.   13 of the 29 counties (45%) with closures are over 40% African-American.  The state of Alabama better prepare to defend these actions in court, because the lawsuits here right themselves.

Looking at the Florida Redistricting Maps offered by the Plaintiffs

When Judge Terry Lewis convenes his redistricting trial later this month, there will be 7 different maps to examine; three from the legislature and four from the plaintiffs.  This article will cover some key differences in the maps being proposed.  The specific focus is going to be on the maps submitted by the plaintiffs and how they compare to the Legislature’s proposals.

Debate Between the House and Senate

The differences between the House and Senate’s final maps are minimal.  Debate centers in the Orlando area and the Tampa Bay region.

Debated Areas

The failure of the legislature to come to an agreement revolves around the mistrust and animosity between the two chambers.  It is truly amazing so little an area derailed the whole process.

  • You can view the House map here
  • You can view the Senate map here
  • The Senate actually proposed another potential map.  Which can be seen here

The Senate compromise map is just bizarre.  The map destroys the rural nature of CD17, stretches CD9 from Orlando to Glades County, and makes other moves that made little sense.  I don’t see any scenario where it becomes law.

The Senate’s initial map is superior in my eyes.  I believe the Senate better handled Hispanic voters in central Florida.  I wrote about this issue in Politico Florida. The key issue is that Hispanics are being unnecessarily split between CD10 and CD9. The Senate map still has some split but it does a better job of keeping Hispanics concentrated in CD9 where they can influence the primary. I wrote about how the Senate map fixes this issue in a separate Politico Florida article

While I think the Senate map is better, I believe something closer to the House map will likely be chosen.  The reasoning is fairly simple, the House map is largely drawn by staff with little legislative input (supposedly).  Judge Lewis may find that argument appealing.  It appears the plaintiffs likewise think the House map is likely to be chosen.  All of their proposed amendments make changes to the House map while ignoring the Senate map.

A key point of debate for the plaintiffs are Districts 26 and 27.  During the redistricting session, the heavily Democratic area of Perrine (an African-American community) was moved from CD26 to CD27.  This made CD26 less Democratic while the 27th, held by Republican Ileana Ros-Lehtinen, became bluer.  Since Ros-Lehtinen is always safe for re-election, and CD26 Congressman Carlos Curbelo is more vulnerable, questions of gerrymandering arose.  I also wrote about this issue in Politico Florida.  Every plaintiff map moves Perrine back into CD26.

I’m going to compare each of these plaintiff maps to the House map one by one.

House Map Compared to First Coalition Map

The coalition plaintiffs (League of Women Voters, Council of La Raza, and Common Cause) propose three different maps. The first map makes the most changes to the House proposal.

Lawsuit House and League 1

The proposal makes changes to several districts across South Florida.

District 20

District 20, the African-American seat held by Alcee Hastings, sees some key changes.  The district loses an appendage it had in Palm Beach, loses part of its central-Broward appendage (both contain African-Americans), but then picks up eastern Miramar from the 25th and 24th.

Lawsuit House and League 1 CD20

The district also sheds less-African American precincts on its boarder, losing much of Sunrise.  The district retains its heavily Democratic nature and would still remain safe for Hastings.  The African-American VAP remains around 50%.  The connection to Miramar is open Everglades land that is unpopulated.

District 21

District 21 sees very small changes.  It takes in the appendage from CD20 and trades some other precincts to even the population out.  The district remains Democratic.

Lawsuit House and League 1 CD21

District 22

District 22 has some modest changes, but the overall nature of the district remains the same.  The district takes in some African-Americans in the Central Florida appendage and loses the city of Plantation.

Lawsuit House and League 1 CD22

The district becomes slightly more Democratic thanks to taking in African-American voters and losing a modestly-Democratic city.

District 23

The 23rd district takes in Plantation and Sunrise while losing coastal precincts further south.

Lawsuit House and League 1 CD23

The district remains solidly Democratic.

District 24

District 24, the African-American seat held by Fredrica Wilson, loses some African-American voters in Miramar largely to the 20th.

Lawsuit House and League 1 CD24

The district goes from 54% Black VAP to 50%.  The district still remains heavily Democratic and Wilson’s only worry would be a primary.

District 25

The 25th trades precincts with the 20th, 24th and 27th.  The biggest changes are made to allow the 20th to stretch into Miramar in South Broward, while it regains population by taking precincts from the 27th further into Miami-Dade.

Lawsuit House and League 1 CD25

The District’s loss of heavily Democratic precincts, especially those in Miramar, take the district from 48% Obama down to 44%.  The district becomes a less attractive Democratic target under this map.

District 26

District 26th takes in Perrine and several south-Dade precincts, becoming much more Democratic in the process.

Lawsuit House and League 1 CD26

The precincts taken in are Democratic while the ones it loses are more divided.  This gives the district a more solidly Democratic footing compared to the House map.

District 27

The 27th sees some major changes.  It loses Perrine and several southern precincts to the 26th and loses some areas to the 25th; meanwhile it takes in areas the 23rd lost on the coast.

Lawsuit House and League 1 CD27

The district becomes less Democratic, dropping around 1.4% in the presidential race but only dropping a 10th of one percent in the governor’s race.  Under the House map, Obama got 54.3% to 51.5%, a gap of 2.8.  The gap under these lines is down to 1.6.  Many of precincts the district losses, especially around Hialeah, saw much larger drops between Obama and Crist.  The removal of those precincts reduced the overall gap between the two Democratic candidates.    The district will continue to be out of Democratic hands until it becomes open.

House Map Compared to Second Coalition Map

The second map by the plaintiffs is much simpler.  Changes are reduced to Districts 26 and 27.

Lawsuit House and League 2

The proposal moves heavily-Democratic, African-American Perrine back into the 26th while taking Hispanic precincts further North and putting them in the 27th.

Lawsuit House and League 2 CD2627

The trading of the two areas, one heavily Democratic and the other leaning Republican, makes the 26th more Democratic than under the House map while the 27th becomes more Republican; having narrowly voted for Scott in 2014.

House Map Compared to Third Coalition Map

The third map by the coalition is an alternative method of giving Perrine back to the 26th.  The map just trades out other areas to even the population.

Lawsuit House and League 3

Lawsuit House and League 3 CD2627

The makeup of the districts is essentially the same as under the 2nd coalition map.  The 26th becomes more Democratic while the 27th becomes more Republican.

House Map Compared to ROMO Map

The final plaintiff map comes from a group of Democratic-leaning voters.  Their map is commonly referred to as the ROMO map.  This proposal also focuses on South Florida, making changes to the 26th and 27th while also maintaining districts 21 and 22 as parallel, rather than stacked.

Lawsuit House and Romo

Regarding Districts 21 and 22, the proposal would re-instate the North-South coastal districts and the North-South inland districts.  This would not change the partisan nature of the districts too much; both remaining Democratic.  The proposal would be a sigh of relief for Lois Frankel and Ted Duetch, who would no longer be drawn into the same district.

Meanwhile, down in Miami-Dade, several precincts get traded around.

Lawsuit House and Romo CD2627

The biggest change, again, is Perrine getting moved back into the 26th. This makes the 26th more Democratic but the 27th becomes more Republican than under any other proposed map.


It appears that most of the debate is focused on South Florida. Districts 26 and 27 still remain a key point of debate and I imagine the issue will be flushed out in good detail during the trial. The different plaintiff maps only nudge the partisan makeup of the congressional lines compared to the House or Senate plans.
The chart below compares presidential and gubernatorial numbers across all different plans.

Redistricting States Plantiffs

While all plaintiff maps make the 26th more Democratic, some of them reduce the Democratic edge in the 27th.  The changes to 27 are a non-factor now due to Ros-Lehtinen’s popularity, but when the seat is open, it will be a different story.  Overall though, both the 26th and 27th remain Obama districts under any proposed map.  The changes proposed, with the exception of the Coalition’s first proposal, are fairly minor and concentrated in small, dense areas.  While the House and Senate may have a bigger court debate over Hillsborough and Orange County, it appears the plaintiffs will largely be arguing over a handful of precincts in Miami-Dade.  However, as the different numbers show, in super-dense South Florida, a handful of precincts can dramatically alter results.


The Complete Breakdown of Florida’s Proposed Congressional Districts

The Florida legislature released its base map heading into next week’s special session on redistricting.  The map makes changes to 22 of the 27 districts.  Some changes are small while others are very large.  Many voters will find themselves in brand new districts if the base map becomes law.  The map below shows areas that will change districts and those that will remain the same.

District Changes

As the map shows, many folks (those in the red) will find themselves in new Congressional districts.  Even those who remain in the same district will likely see the nature of their district change.

The partisan effects on these districts slightly benefits Democrats.  Gwen Graham’s district gets carved apart and CD2 becomes very Republican, netting the GOP a seat.  However, district 13 and 10, held by Jolly and Webster, become very Democratic and are likely pickups for Team Blue.  This results in a net gain of one district for Democrats.

The Florida redistricting site offers the estimated Presidential vote from 2012 for the new districts (as well as the old).  I have attached the Presidential totals for the current districts and the new districts.  However, it should be noted that due to a large number of precinct splits, exact calculations are impossible. I included another Obama % (Alternate) which reflects calculations I made based on how I assigned the split precincts.  In many districts the difference is minimal, in others it is higher (those with more split precincts).  Overall the fluctuation is never more than a few 10ths of a percent.  I also personally calculated estimated Gubernatorial numbers from 2014.  Again, due to split precincts, these are estimates and they also factor in all third party and write-in votes.  The gubernatorial numbers are rough estimates that could fluctuate (thanks to split precincts) by a few 10ths of a percent with a deeper dive.  This is meant to be a general view

Totals Stats

Democrats see improvements in some districts and losses in others.  Many districts saw very minimal partisan change.  Democrats improve their showings in CD7 (Mica) and CD27 (Ros-Lehtinen) while losing votes in CD14 (Castor) and CD9 (Grayson). They have a few more competitive seats with this map, but the delegation remains in GOP control.

I have gone through each district and done a write-up on how they have changed.  Several proposed districts look good.  Others have major problems.

District 1

Jeff Miller’s solidly-Republican district was one of just a handful to see zero changes. The district is based in the western end of the Panhandle and will remain safe for the longtime incumbent.

District 2 

District 2 undergoes massive changes.  In order to create a Jacksonville to Tallahassee district, CD2 loses Gadsden, half of Leon, most of Jefferson, and the part of Madison it held.  In order to balance the population, it takes in 7 more counties:  Suwannee, Lafayette, southern Columbia, Gilchrist, Dixie, Levy, and western Marion.

2nd District

This new district is heavily Republican, giving Obama around 34% of the vote.  Even Bill Nelson, one of the few remaining Democrats to do well in the Panhandle, only received 44%. Gwen Graham resides in this district but has little chance of winning it if she runs.  She faces the double-whammy of losing Democratic areas and adding many new, heavily-conservative areas that have no reason to support her.

The largest population bases are Leon and Bay.  Bay has the advantage over Leon, holding more registered Republicans than Leon.  Bay has 31% of the Republicans in the seat to Leon’s 18%.  The rural counties would need to have a unified force to elect a Representative of their choosing.

CD2 Stats

There is a major problem with splitting Tallahassee, but I will address that in district 5′s analysis.

District 3

Ted Yoho’s district undergoes massive changes. Yoho loses many of the rural counties that propelled him to his primary upset in 2012.  Meanwhile, Yoho takes in the rest of Alachua and Clay, Putnam, and northern Marion.

3rd District

Compare the map of what he loses to his primary win in 2012 over Cliff Stearns.  One of the reasons Stearns lost was due to 2012′s redistricting and losing parts of his base (Ocala).  Ironically, Yoho finds himself in a similar position.

CD3 Republican Primary

Yoho loses some of Marion that didn’t back him, retains Clay, and loses many bases of support.  In such a GOP district, Yoho is only vulnerable to a primary, if a serious one emerges.  Yoho could hold on and entrench himself.  However, the establishment has no love for him and a strong challenger could make it interesting.  The three biggest GOP bases are Clay (40% of the district’s Republicans), Alachua (25%) and Marion (21%).

District 4

Ander Crenshaw’s district loses Baker and part of Duval to the new 5th district and takes in northern St. Johns to balance the population.  The district should remain safe for Crenshaw and any Republican.  The addition of St. Johns will give a Republican from that area an opening to run in the future.

4th District

This district remains one of the most Republican in the state.

District 5

The legislature followed the court order by drawing Corrine Brown’s 5th Congressional district east-west.  The district goes from Duval, takes Baker, Hamilton, Madison, northern Jefferson, half of Leon, and Gadsden.  The district is 45% African-American.  Meanwhile the district loses everything south of Duval.

5th District

The district’s main problem is that it splits Leon County.  As I demonstrated here, the district can retain African-American status and keep Leon whole.  Under this map Tallahassee is carved to pieces as African-American areas and some white neighbors are added into the 5th district.  The Leon County map below shows the areas that will be in CD5 (blue) and CD2 (red).


The two maps below show the district taking in African-American areas in Tallahassee and Jacksonville. Many precincts are split in the process.

5th District black5th District black jax

Corrine Brown will be favored in this district.  The seat is heavily Democratic and 66% of the registered Democrats are African-American.  55% of the Democrats in the district come from Duval, which is actually less than in the League of Women’s Voters’ proposed map.  Leon, down at 25% of Democratic voters, has little chance of influencing the district.

5th Stats

Many members of the Leon delegation have expressed desire to keep Leon whole in CD5.  This will be a debating point when the special session begins.

District 6

District 6, left open due to DeSantis running for US Senate, undergoes several changes.  The district loses northern St. Johns to CD4 and the bit of Putnam it had to CD3.  The district then takes in parts of Marion and Volusia to balance the population.

6th District

The district remains Republican favored, albeit less so than under its current lines.  Under the current boundaries, St Johns and Volusia had the same number of GOP voters.  However, with Volusia being made whole and St. Johns being split, the Republican base shifts more to Volusia.

District 7

Mica’s 7th district undertakes some very surprising changes.  The district loses southern Volusia to CD6, losing Deltona.  However, it is the parts of Orange that it takes in that are surprising.

7th District

The district loses some parts of Orange on the west but then moves south to take in a good deal of eastern Orlando.  Many of the precincts it absorbs have significant Hispanic populations and currently reside in CD9, a Hispanic access seat.  These changes makes the district much more Democratic, going from 47% Obama to over 49%.  However, the taking in of Hispanic voters may become an issue in the special session and under court order.  Mica’s district becomes bluer, but he is still a favorite for re-election.

District 8

District 8 had no changes and should remain safe for Bill Posey.

District 9

There is no way around this.  There are many problems with District 9.  The district, being vacated by Alan Grayson, loses many of its Hispanic voters, falling to 30% Hispanic (32% if you include Hispanic Blacks in the figure).  When the district was drawn in 2012, it was 39% Hispanic and 41% if Hispanic Blacks were included. The district becomes much whiter thanks to losing Hispanics in Orange and gaining whiter areas form Polk.

9th District

The district becomes less Democratic, falling from 60% Obama to 55.9%.  What is especially notable is that in the midterms Crist fell to 48.1% (though he won with Scott getting 46%).  Alan Grayson would not be able to hold the seat in a midterm.

When looking at the Hispanic registration in the area (which tends to under-perform total population) compared to the new boundary, it is clear many Hispanics could be added into the seat.

9th Hispanic

The boundary seems to go out of its way to take in white areas of eastern Polk while losing heavily Hispanic areas in south Orange County.  The Hispanic population is unnecessarily split and the seat is no longer an access seat for the community.  This must be addressed in the special session.  Otherwise, a lawsuit will follow and The Supreme Court may not approve the boundary.

District 10

If one Republican Congressperson was going to be sacrificed, I predicted it would be Webster.  Webster, whose district is currently around 47% Obama, angered the GOP establishment with his challenge to Boehner for the Speakership.  With African-Americans from CD5 needing to go somewhere, Webster seemed like the likely candidate.  The extent to which he was thrown overboard though is staggering.  The district takes in the African-American voters of CD5 and many Hispanics of CD9.  Meanwhile, it loses Republican-friendly Lake.  The district becomes heavily Democratic, giving Obama 60% of the vote.

10th District

The district, which used to be over 60% voting-age population white, is now 44% white, 25% Black, and 21% Hispanic.  For someone like Val Demings, the district is a perfect opportunity.  Many other Democrats may look at the seat now as well.  However, as I mentioned with CD9, there is an issue with the Hispanic population.  CD10 may need to give Hispanics back to CD9 and take in other areas.  The end result should be a Democratic district, but less than this proposal.  Because of possible changes to CD9, CD10′s boundaries may likely change.  Regardless, Webster is in deep trouble.

District 11

Richard Nugent’s district undergoes some changes as it loses Ocala to CD3 and picks up parts of Lake from CD10.  The district remains reliably Republican.

11th District

The district should not present any real problems from Nugent, who hails from Hernando County.

District 12

District 12 also undergoes some slight changes, taking in parts of Pinellas from CD13 and evens out the population by losing parts of Hillsborough.

12th District

The district is still modestly Republican and Gus Bilirakis will have no trouble with re-election.

District 13

David Jolly saw the writing on the wall after the Supreme Court’s redistricting decision.  With CD13 being forced to take in African-American St. Petersburg, there was little chance the district wouldn’t become much more Democratic.  As predicted, the district took in the African-American areas and lost some of its northern precincts, largely retaining Clearwater on its Northern end.

13th District

With Charlie Crist publicly expressing a desire to run for the seat, the district is a likely Democratic pickup.  With polls showing Crist solid in the primary and general, it is a very good chance Crist will make a comeback with this Congressional run.

District 14

District 14 sees changes as it loses the African-American parts of Pinellas and gains many communities in the North.  The district almost makes Tampa completely whole.  However, the boundaries may run into problems on racial grounds.

14th District

District 14 remains Democratic, albeit less so than before.  However, the real issue for the seat is the changes in its demographic makeup.  The district was originally drawn as a minority access seat:  with 46% of the voting age population being white, 24% black, and 24% Hispanic.  However, under these lines the district is 52% white, 17% black, and 24% Hispanic. Now the African-American share was going to fall due to the loss of South Pinellas.  However, many areas with significant Hispanic and Black populations are left out of the seat while whiter areas are taken in.

14th District Black14th District Hispanic

The district could move east to take in the community of Brandon, and maybe even Plant City, while losing some of the whiter areas in the west (giving them to CD12 or CD15).  In fact, the district 15 seat could take in many of the northern Hillsborough areas and could go further west to take in more of them.  The split to Tampa could be minimized (it still isn’t entirely whole in this proposal).  As two of my proposals show, it is possible to draw the district keeping the white voting age population under 50%.

District 15

District 15 undergoes some major changes.  First it takes in the part of Lake taken from Webster’s district, then it takes parts of Hillsborough around the Brandon areas, gives up the Northern Tampa area, and loses parts of Polk.

15th District

The new district is still modestly Republican. However, Ross will have many new voters.  However, he should be fine for re-election, as only around 40,000 new Republican join the district and 120,000 remain in the area he already represents.  However, with the issues I laid out in District 14, the 15th may need to give up the Brandon area and take in some more of northern Hillsborough.

District 16

Vern Buchanan’s seat is undergoing a surprise change as it loses Southern Sarasota and takes in parts of Hillsborough.

16th District

Buchanan’s district seems to swap areas with CD17, seemingly for no clear reason.  Much of the Hillsborough area gained by 16 is taken from 17, as is eastern Manatee, while 17 takes southern Sarasota County.  Buchanan should be favored in this Republican seat, which includes his base in North Sarasota.  He retains more GOP voters than he takes in.  However, there may be significant debate over the seemingly unneeded split of Sarasota County.

District 17

Tom Rooney’s safe Republican seat doesn’t get any more competitive.  However, some of the changes to the district are questionable.  As discussed with CD16, the splitting of Sarasota seems unnecessary.  The district now includes Southern Sarasota, while giving CD16 a good chunk of Hillsborough.  The district also trades areas of Polk while losing the low-populated region of Southern Osceola.

17th District

The tradeoff of Sarasota and Hillsborough with CD16 is the biggest issue here.  Regardless of any changes, Rooney, based out of low-populated Okeechobee County, has never had a tough time getting re-elected.  However, I believe debate will be had over the splitting of Sarasota.  Regardless, Rooney is very likely safe from any primary.

District 18

District 18, based out of St. Lucie, Martin, and Northern Palm Beach, has no proposed changes.  The seat will be heavily fought over by both parties in the fall.

District 19

District 19, held by Republican Curt Clawson, also sees no changes.  The seat remains safely in the GOP column.

District 20

Alcee Hastings’ district undergoes very minor changes.  The Supreme Court ordered Hendry be made whole in a district, and this map opts to make it whole in CD25.  The population of Hendry is very low and it only takes a few added precincts in Broward to even the population out.

20th District

The district remains 50% African-American (48% if you don’t count Hispanic Blacks) and is heavily Democratic.

District 21

When the Supreme Court ordered Districts 21 and 22, Democratic seats held by Ted Deutch and Lois Frankel respectively, be redrawn in a more compact manner (while maintaining District 20′s African-American majority), the prospect of both members ending up in the same seat was high.  The current districts both split Palm Beach and Broward, running parallel to each other.  The Court advocated, but did not demand, stacked districts.  Indeed, under the proposal submitted, District 21, based out of Palm Beach is stacked on top of District 22, which includes Boca and parts of Broward.  District 21 contains the homes of both Deutch and Frankel.  Frankel lives on the Northern end of the district, while Duetch lives in the eastern neighbors outside Boca in the Southeast end of the district.  The seat retains a great deal of Deutch’s current district, loses the parts that were in Broward, and takes in the northern half of Frankel’s 22nd district.

21st District

Deutch lives very close to what will be the new 22nd district, just miles north of the Broward line.  Frankel, meanwhile, lives much further north and would face trouble running in a district much further south that she does not live in.  Frankel is also a well-known institution in Palm Beach County, with a long career in the State Senate, Mayor of West Palm Beach, and now Congress.  However, 63% of the Democrats in the proposed district are in Deutch’s current seat.  Deutch also commands a very loyal following, especially in the Jewish community, and Frankel cannot easily take his voters to win a primary.  If both ran for the seat, it would be a titanic primary.  While I initially believed Frankel would be heavily favored, it appears Deutch would have an edge due to representing more voters and those voters remaining very loyal to him.

District 22

With District 21 taking in much of the northern end of this current district, 22 ends up based much more out of Broward, taking in the Broward areas Deutch represents now and taking in white areas surrounding CD20.  The district also retains Boca, giving it some ties to Palm Beach County.

22nd District

Both Frankel and Deutch will at least look at running for this seat.  For Deutch, the Broward area he currently represents that will be in this district, has many condo areas with large Jewish populations that would likely be loyal to him.  He would have to deal with the coastal and southern areas that don’t know him, but would likely start off with a moderate advantage.  Frankel could also run here. However, she is likely to have less of a loyal following in Broward than Deutch commands.  If a Broward politician challenged her in the primary, the Broward regions Deutch represents now would be up for grabs and the coastal and southern areas could swing away from her as well.  Frankel will be coming all the way from West Palm Beach and that could become an issue.  It is also worth noting that when Frankel won her primary in 2012 with 61% against Broward Commissioner Kristen Jacobs, they tied in Broward.  A Frankel v Jacobs re-match could yield very different results under the new lines.  There are also several other Broward politicians who could make a play for the seat.  Frankel or Deutch may not get a primary to themselves if they try to run here (and it would be smart to make a decision quickly in order to stop a potential primary).  If neither make the move south, then the seat could be a free-for-all of Democrats.  State Representative Kristen Jacobs may look at the seat.  Meanwhile 3 members of the Broward Commission  represent parts of the district:  Mark Bogen (disclosure, I worked for his campaign in 2014) represents many of the heavily Democratic condos of the district, Stacy Ritter represents the northwestern neighborhoods of Parkland and Coral Springs, and Tim Ryan represents parts of the coast and south.  There will be no shortage of options for Democrats here.

District 23

Debbie Wasserman Schultz’s district is undergoing minor changes. The district loses part of Sunrise to District 22 while taking in a few precincts of Pembroke Pines from CD25.    The district also takes in a small portion of south Ft. Lauderdale.

23rd District

The seat remains safe for Democrats and the DNC Chair.

District 24

No changes were made to this heavily Democratic, majority-African American district.  Congresswoman Fredrica Wilson continues to be a shoe-in for re-election.

District 25

Per the court order to make Hendry whole, district 25 took in the eastern half of Hendry while losing a few precincts in Pembroke Pines to even the population out.

25th District

The district has trended toward Democrats in Presidential years.  The district may get much more competitive if trends continue.  However, the midterms did not go well in the district.

District 26

When the Court ordered the Democratic city of Homestead be made whole (currently split between CD26 and CD27), many, including myself, thought they would give the city to CD27 in order to make the 26th more Republican.  Both seats are held by Republicans, but only Congressman Carlos Curbelo is being seriously challenged.  Curbelo won his seat in 2014 and is a top DCCC target.  District 27 Congresswoman Ileana Ros-Lehtinen is very popular, liberal on social issues, and can afford to take in more Democratic voters. However, the map actually makes Homestead whole in CD26, giving the district more Democrats.  However, the district also loses several heavily African-American precincts further north to even the population out.

26th District

The removal of the African-American precincts makes the district more Republican than it currently is, taking Obama’s % from 53% to 51.6%.  Under current lines, the Crist/Taddeo ticket got 50%, now it has 48.9% (a plurality win with Scott getting 48.3%). Taddeo is running against Curbelo and does have a shot at taking the seat.

It should be noted how…. convenient… it is that the precincts moved to CD27 are so heavily African-American.  While 26 takes all of Homestead, it already had the African-American western side of the city.  The African-Americans it loses in the north are much more Democratic than the rest of Homestead.  The map below shows the African-American pocket that is transferred from CD26 to CD27.  The new 27th boundary loops around and takes in the African-American voters.

26th District Black

District 27 does already have some African-American voters.  However, the move of the pocket in the North seems like a clearly partisan move.

District 27

The changes to district 27 are already highlighted in district 26.  Thanks to taking in the heavily African-American areas north of Homestead, the district becomes more Democratic.  However, Congresswoman Ros-Lehtinen remains safe due to her personal appeal.

27th District

The second she retires, however, this seat will be a prime Democratic target.


With special session beginning next week, there is still a long way to go before anything is settled.  Many districts are likely to be debated and challenged and no doubt some changes will take place.  However, from this base map we can see the trend.  Democrats will benefit from this map or any variation of it.  However, the benefits are still limited and more work will need to be done for the party to win control of the Congressional delegation.

Leon County Must be Kept Whole in any Congressional District

In today’s Tallahassee Democrat article, DNC member Jon Ausman and myself advocated for keeping Leon County whole in any new congressional district drawn in next month’s Special Session.  The Supreme Court ordered that the current 5th Congressional district, stretching from Jacksonville to Orlando, be eliminated and a new east-west district be drawn.  The court cited the proposed districts by the League of Women Voters as an acceptable option.  The League’s proposed map splits Leon County between two districts, thus diluting the power of the County and Tallahassee.


Under these lines, or any similar variation, Leon County will get short-changed.

This proposed map should be a concern for Tallahassee voters, Democrat or Republican.  Depending on where in the city you live, you will find yourself either in the 5th Congressional district, the Jacksonville to Tallahassee proposal, or the new 2nd congressional district, a collection of rural counties that goes from Bay to Levy County.

In the newly proposed solidly Democratic District 5, Jacksonville will make up around 58% of the registered Democrats with only 20% coming from Leon.  Leon County residents in District 5 will be stuck with whoever wins the primary, likely a Jacksonville candidate. This proposed district would have given Obama an overwhelming 61% of the vote; essentially meaning it would be near impossible for a Republican to ever win District 5.

For Leon County residents living in Solidly Republican District 2, the chance of true Representation is not much better. Under the newly proposed boundaries, Bay County will make up around 31% of Registered Republicans, with Leon around 20%. The rest of the district will be made of low populated but highly conservative counties. Democrats will have no chance at winning the seat, which would have given Obama only 35% of the vote.

Splitting Leon County is not necessary.  A district can be drawn from Jacksonville to Tallahassee that keeps Leon County whole while still creating a district likely to elect and African-American, as mandated by the court.  The district below keeps Leon whole and maintains a 42% Black Voting-Age Population.


Under this district, Leon County would have 42.9% of registered voters while Jacksonville would have 38.3%.  In a Democratic primary, 38% would come from Leon, with 41% from Jacksonville.  The primary is very likely to decide the Represenative in this 63% Obama district.  The Democratic voters are also 60% African-American, making it likely an African-American Democrat will represent the seat.  What this district does is give balance between the two major cities, ensuring any Representative will have to cater to both areas to ensure re-election in a primary.  For Leon Republicans, this is a good thing as well.  While they may not get a GOP congressperson, they at least will have a Congressperson who must cater to the needs of their county.

I encourage you to read Jon Ausman’s op-ed in the Tallahassee Democrat, further making the case for keeping Leon County whole.  Keeping Leon whole creates balance between the east and the west, something that will benefit both cities by ensuring a Representative that is always eager to make both areas happy.  

Update:  Base Map out

The base map is out and it doesn’t look good for Tallahassee.  The proposed boundary is similar to the League of Women Voters map; splitting Tallahassee.  The map below shows the Tallahassee area precincts and which precinct goes were.  Red is the new CD2, tied to rural areas, and blue is CD5, tied to Jacksonville.


Greek Referendum Results Outpaced Anti-Austerity Parties

Greece held its referendum on whether or not to accept the terms of its bailout from the European Commission, International Monetary Fund, and the European Central Bank.  For years, Greece has gotten billions in bailout funds with conditions heavily rooted in austerity: deep social cuts and tax/fee increases. For the last several years, Greece has been forced to cut public works, pensions, and social programs.  Meanwhile, Greece has been forced to raise taxes and fees on working and middle class voters; while proposed cooperate tax increases were not allowed by the IMF.  The rigid conditions placed on Greece have fostered a sense of anti-European and anti-austerity sentiment in the country.  Citizens have complained about the major European powers, notably Germany, dictating the policy conditions that Greece has been forced to adhere to in exchange for bailout funds.   The argument has been that such dictations are anti-democratic.

The economic conditions in Greece continue to be extremely poor thanks to the forced austerity measures. Unemployment hovers around 25%, poverty rates are high, and businesses have been shuttering constantly.  The conditions have lead to a collapse of the old party system in Greece.  For decades, the center-right party, New Democracy, and the center-left party, PASOK, have dominated elections.  However, with both parties accepting the rigid bailout terms, their combined support began to collapse a few years ago.


As the graph shows, both parties amounted to over 80% of the votes cast until the chaotic 2012 elections.  2012 saw the rise of anti-austerity Parties that seized on the economic conditions of Greece.  The socialist party, SYRIZA, came in second in the June 2012 snap election.  Meanwhile, Golden Dawn, a scary pro-Nazi party, began to make gains by seizing on discontent and anti-immigrant sentiment.

In the January 2015 elections, SYRIZA gained 36% of the vote, the largest of any party, and was able to form a Coalition Government with another anti-Austerity Party, ANEL.  SYRIZA did so well they secured 149 of the 300 seats, just two short of an absolute majority.  This made Alexis Tsipras, a fierce opponent of austerity, Prime Minister.

Months of negotiations have led to stalemate between Greece and the European Powers.  Europe has insisted on more austerity measures as a condition for more bailout funds.  Greece just defaulted on its IMF payment last weak, stating simply that they did not have the money.   Prime Minster Tsipras decided to put the conditions laid out by Europe up for a vote of the people.  Tsipras vowed to accept the results, but that he would resign if the YES won, stating he would not help implement austerity measures.  When voters went to the polls, they voted to reject the conditions with over 60% of the vote.

The map below shows the referendum results.  Every electoral division rejected the the bailout conditions.

Greece Votes2

Next to the referendum I showed the results from the parliament elections that swept SYRIZA to power.  I combined all the anti-austerity parties together:  SYRIZA, Golden Dawn, KKE, and ANEL.  Note this only includes parties that won seats in parliament.

The scatter-plot below compares the NO vote on the referendum to the support for the anti-austerity parties in the last election


There is a notable statistical correlation between the anti-austerity parties and the NO vote on the referendum.  The NO vote outpaced the anti-Austerity parties in every district.

I also compared the NO vote with the support for the SYRIZA party in the January elections.  The statistical correlation was slightly weaker, but still there.


Meanwhile, the referendum had a strong statistical correlation with support for the right-of-center New Democracy party, which is the second largest party in Greece.  In this case, the stronger New Democracy did in the last election, the worse NO did in the referendum.



This shows that many New Democracy supports favored the bailout agreements, either due to conservative economic views or a desire to stay in the Eurozone.

No other parties, most of them getting small percentages in the last elections, had a strong statistical relationship with the referendum results.

A good deal can be said for what may happen next.  Greece may leave the Eurozone.  If there is a deal to be worked out between Greece and the European banks/powers, those institutions should look carefully at these results.  Many of the European power players have complained about Tsipras and how hard it has been to negotiate with him.  However, the referendum results show the Greek people stand with their Prime Minister and the anti-austerity forces.  If the rest of Europe wants to work out a deal, they need to stop using the Prime Minister as a scapegoat and realize the Greek voters will not accept stringent austerity anymore.



Why Can’t TV Shows Create an Electoral Map that makes Sense

The Season 4 finale of the hit HBO show, Veep, saw President Selina Meyer learn her fate as she sought to get elected to a full term as President.  The electoral map that resulted from the episode was a little odd for anyone who follows Presidential politics. States were won by parties you would never expect to see in real life. The whole episode got me thinking about the times we saw election results in TV shows that just did not fit in line with reality.  Moments that make people like me get angry with nerd rage.

Comic Book Guy

Hence I have decided to do a post on the odd electoral maps we have seen in big hit TV shows.  The question for me is, why do these maps have to be so unrealistic?


Lets start off with the most recent offender, Veep.  In the finale, Meyer was facing off against a presumably-Republican US Senator in what was considered a tight race. While Meyer is never officially identified as a Democrat, her agenda throughout the series indicates she leans left of center.  As the finale begins, state’s start getting called. A close observation of the episode gives us a good idea of which states were called for which candidate.  Some states calls like Pennsylvania, Colorado, or Vermont are uttered by cast members, while views of the electoral map on the TVs help fill in the other gaps.  The episode ends with the electoral college in a direct tie.  Based on the calls made in the show and the maps seen on the TVs, this is the most likely electoral map for Meyer.

2014 Veep

Anyone who follows Presidential politics can see some problems with this map. Minnesota is red for the first time since 1972.  Delaware is red for the first time since 1988.   Meyer is also losing notably Democratic states like Nevada and Michigan, yet winning key swing states like Colorado, Florida and Virginia.  A scenario where Delaware and Minnesota go red is indicative of a Republican landslide in our modern political world.  Florida going blue also seems unlikely since Amy, Selina’s old campaign manager, sings that turnout is low in Broward County, a huge Democratic base of votes for any Democrat running in the state.  Low Broward turnout tends to mean Florida going red.

If the writers wanted to have a tied election, there was a much more plausible map. Meyer could have lost much of the Midwest but retained other swing and Democratic states.

2014 Veep Alternate

In this scenario, Meyer loses most of the rust belt and the farmland areas thanks to dips in support from the heavily white population in those areas.  However, Meyer wins Florida and Virginia due to heavy GOTV efforts with Florida’s minority population and in Virginia’s Northern suburbs.  Virginia is continuing to become more Democratic as its population shifts and Florida often comes down to turnout in key Democratic counties. Minnesota and Delaware go back to blue and the map reflects something a but more in line with reality.

The West Wing

Veep isn’t the only show with odd electoral maps.  The West Wing was a great offender on this front as well.  While only one election in the show featured an official electoral map (the 2006 race), from conversation of characters, its possible to recreate the electoral maps of Bartlet’s two elections.

First is the 1998 election that Bartlet won to become President.  This happens a year before the show begins.  From conversations between characters, we know Bartlet got 48% of the vote, but got more votes than an unnamed Republican opponent.  Bartlet also got 303 electoral votes.  Different conversations and quips note that Bartlet won states like Indiana, but lost states like Maine.  It is also implied Bartlet won Georgia and Montana.  Most other state’s mentioned are fairly normal. However, to get to 303 votes based on the states mentioned, the map looks something like this.

1998 West Wing

Its hard to see a universe where Bartlet loses Minnesota (similar to Veep) but wins Florida, Georgia, Indiana and Montana.  Bartlet losing Maine also seems to be because they don’t seem to like him very much, by his own admission.  Bartlet is implied to have won Indiana thanks to the Democratic Governor’s campaigning for him, while Hoynes delivered Georgia, West Virginia, and probably Florida.  I tried creating a map with blue Minnesota but could not get to 303 based of states staying they way they were called in the show.  This map is the most subject to speculation.  However, other fans have come up with the exact same map (you can find discussions of it online).

While the 1998 map has issues, the 2002 map gets even weirder.  Bartlet wins re-election thanks to crushing his opponent, the Governor of Florida Robert Ritchie, in the one debate of the election.  Bartlet’s close re-election turns into a blowout following the debate and he wins by a 55% to 44% margin.  The episode following the polls closing sees characters rattling off the different states’s Bartlet is winning, mostly focusing on red states they never expected to carry.  Based on the states mentioned, this appears to be the most likely electoral map.

2002 West Wing

There is some variation to this, but the consensus appears to be a 39 state win.  Bartet is said to have swept the plains states, the Great Lake States, won much of the upper south, won Louisiana, and presumably won every swing state.  For no logical reason, New Hampshire, Bartlet’s home state, is close.  Despite the big win, Bartlet losses the deep south and Texas (his VPs home state).  It is possible the map is different, but Bartlet sweeping the plain states (including Nebraska and the Dakotas, as well as winning the states travelling down the Mississippi river) is cannon.  The map honestly seems too blue for a 9% win.  When Clinton beat Dole  by 9% in 1996, he did not win the plains states, which are heavily Republican (Montana is a bit less red).  Indiana is also much more blue in the West Wing universe than in real life, despite Toby Ziegler once commenting that Indiana would vote for any Republican.

One key piece of continuity issues occurs during the 2006 election when it is commented that North Dakota hasn’t voted Democrat in 40 years.   Yet in season 4 it is expressly states Bartlet won by the Dakotas.

The last West Wing map is the 2006 election that saw Democrat Matt Santos beat Arnold Vinick in a narrow election.  The map comes directly from the show, making it the least disputable map in this article.  However, the map still has some very odd results.

2006 West Wing


So lets address California and Texas first.  In the show, Arnold Vinick is the Republican Senator from California.  He is pro-choice, socially moderate, pro-environment, and fiscally conservative.  He is by far the most liberal Republican to run for President and have the nomination since Teddy Roosevelt.  Matt Santos is the young, energetic Congressman from Texas.  Vinick’s liberalism and personality originally make the election his to lose and he has a commanding leading the electoral college.  However, a near meltdown of a nuclear plant in California (and Vinick’s support for nuclear power and his push for the plant to be opened decades back) make the race a tie.  Vinick still wins his home state of California narrowly.  It is implied he would have dominated it without the incident.  This is questionable on its face.  Even if a Republican Senator existed in California, the states blue lean would still make it close from day one.  Unlike Governors, Senators are less know and it is questionable that Vinick had such a death-grip on the state.  Santos, meanwhile, wins Texas despite only representing a blue district out of Houston.  Texas didn’t vote blue in Bartlet’s blowout, I question it voting blue now.

In addition, Vinick wins Iowa despite coming out against ethenol subsidies during the primaries. Vinick also wins Vermont and Maine. Maine is onr thing, but Vermont, I don’t care how liberal Vinick is, seems unlikely.  Meanwhile, Santos still wins New Hampshire while losing Vermont, an odd result.  The weirdest result is Santos wins South Carolina for no real reason other than the state having a Democratic Governor.  Santos loses Florida because Latino’s aren’t showing up like expected, which does seem unlikely with the first Latino Presidential Candidate and the campaign’s focus on the state.  The electoral map for 2006 has many issues besides just California and Texas.


These are the most prime examples of electoral maps being used in major political TV shows.  It is strange to see shows with such a political bent and attracting political audiences using election results.  These are the shows were a higher percentage of the audience would be able to look at the map and call BS right off the bat.  One can’t help but wonder if the maps are specifically meant to troll the more political side of the audience.  If so, then I tip my hat to them.  After all, they got me to write this article.


The Disappearing Swing Counties of Florida

Over the last 20 years, Florida’s reputation as a swing state has endured.  The state’s population continues to grow at a fast rate, recently propelling Florida past New York, making the state the 3rd largest in the nation.  Yet while other states have seen population shifts effect the political leanings of the state; Florida has generally remained a swing state in Presidential politics since the 1990s.  Migration into Florida has come from the Midwest, South, and Northeast; while the population of those born outside the US has risen dramatically in recent decades.  Many of the new residents from Northern States have been Republicans while others have been Democrats.  Many Northern migrants have been more Republican suburbanites, while others have been more urban/liberals.  Meanwhile the growing out-of-US population has also had its share of Democrats and Republicans.  A growing Puerto Rican population is heavily Democratic, while the Cuban immigrants have been historically Republican, with the exception of the millennial generation.  Overall, Florida’s competitiveness during Presidential elections has held from 1992-2012.  The state still leans more Republican than the nation, and the GOP controls the state Government thanks to a money advantage and 2 decades of gerrymandering.  Democrats have actually seen a reversal of fortune from the 1980s, where the GOP dominated Florida in the Presidency but Democrats maintained control of the state Government (for the most part).  Since the 1990s, the GOP has taken control of the legislature and cabinet.  However, Democrats have still been competitive in Presidential elections, winning 3 (or 4) of the last 6, and the last two Gubernatorial Elections were decided by around 1%.

The population growth of just the last 20 years has transformed the state but not caused Florida to fall off the swing state list.  However, shifting populations and shifts in how different groups vote has effect the ways different counties in the state vote.  Shifting electoral dynamics, from urban areas becoming more Democratic or rural voters growing more Republican, has also caused a major shift in how individual counties in Florida vote.  Over the last 20 years, Florida Democrats have increased their support in urban and diverse counties, while at the same time seeing dramatic falloffs in support in North Florida and the rural regions of the state.  All of these shifts have balanced out as both parties maintain a shot at the state’s electoral votes.  To highlight this shift in support among the counties, I looked at the last 6 Presidential elections, all of which were competitive, to see how counties voted and by what margin.

It should be noted before delving into each election, that while the state has seen many shifts in population, there has been only marginal shifts in terms of each county’s share of the vote in Presidential elections.  The map below shows how each county’s share of the statewide vote has shifted from 1992-2012.

Shift in Share

As the map shows, only a few counties saw there share rise or fall by more than 1%.  Orange has seen the greatest growth, 1.04%, while Pinellas saw its share fall by 2.5%.  Most counties saw their share shift by less than 0.5%.  So when looking at maps back to 1992, yes some of these counties were worth more or less than they would be now, but the general dynamic remains the same.  The more dramatic shifts, the huge growth of the Southeast and the I-4, began before the 1990s.  All the major counties continue to see massive population increases, they just aren’t dramatically outpacing their neighbors.

With that said, lets look at the last six Presidential elections and how the counties voted.

Presidential Elections 1992-2012

In looking at each Presidential election, I examined the margin that the county was one by, rather than just the percent a candidate won. In each election I provided a map of the win margin and a bar graph showing how many counties fell within different margins.  Counties were categorized by win margins of 0-5%, 5-10%. 10-15%, 15-20%, and 20% of more.  Counties that feel within the 0-5% range are referred to as swing counties (because of their narrow wins) in the election summaries below.  You can select the maps to see larger versions.

Bill Clinton narrowly lost Florida to Bush in 1992.  Clinton had not counted on winning the state and hence invested fewer resources into it.  Nevertheless, Clinton won 22 counties, several of them rural, and narrowly trailed in several others.  19 counties were won by 5% or less and only a handful were blow-outs.

1992 Margin1992Swing

It should be noted how few heavily GOP counties there were in this election despite Bush’s win.  Many counties fell within swing or modest win zones.  A shift of a few points statewide could have thrown several more counties into either the Bush or Clinton camps.  Clinton did well in the Southeast and many traditional Democratic counties we have today, but the Orlando region was notably still in the red column.

In Clinton’s 1996 blow-out of Bob Dole, the President won 32 counties,and 20 counties were decided by 5% or less.  Clinton made several modestly GOP counties closer compared to 2012, while at the same time increasing his support in counties he won the first time, keeping the swing county number close to the same.

1996 Margin1996swing

The 1996 bar graph shows even more counties clustered toward the middle, giving swing or modest wins.  Only a handful of counties were heavily in the GOP or Democratic camp,  Clinton’s win marked the last strong showing for a Democrat in the rural counties in a Presidential Election.  Orange County, home to Orlando, is notably still red.  Orange would become more blue as the share of non-white voters grew in the County.

Al Gore won, I mean lost, Florida by the narrowest margin imaginable.  On the county level, the number of swing counties fell to 10 as several swing counties became more GOP entrenched.  The number of counties backing the GOP by 20% or more jumped from 9 to 20 as rural counties, most in North Florida, made hard-right shifts.  Leon, meanwhile, fell into the 20%+ margin for Democrats and Orange became a blue county as population growth, especially among Hispanics, changed the county’s makeup.  As urban counties like Dade, Orange, and Palm Beach continued to grow and become more diverse, they became more Democratic.  Gore did worse than Clinton’s 1996 landslide win in some of these urban counties, but did better than his 1992 showing.

2000 Margin2000Swing

This election marked the beginning of a clear trend, the urban counties being firmly Democratic, and the rural counties going more Republican.  Gore’s improvements over Clinton’s 1992 election (where the margin was close in Florida) showed a Democratic strength increase in those areas.

Bush won Florida by 5% in 2004, narrowly falling short of Clinton’s 1996 showing.  Bush managed to make Kerry perform worse than Gore in most urban counties and dramatically increased the number of 20%+ GOP margin counties.  Bush continued to trend of the GOP locking up the small rural regions.

2004 Margin2004Swing

Kerry’s loss was a stumbling block for Democrats, as Bush continued the trend of the rural counties going more Republican.  Meanwhile many urban counties saw weakened Democratic margins.  Notably though, Miami-Dade’s Democratic margin of 6.2% was unchanged from 2000, despite Bush’s Hispanic appeal in 2004. Bush also improved the Republican margin with suburban counties on the west coast, increasing already solid GOP margins.  Several counties that fell within 5% in 2000 (either for Gore or Bush) fell into the 5-10% camp for Bush, reducing the number of swing counties to 7.

Obama won Florida by nearly 3% in 2008.  He did this by moving modest GOP counties into the swing region, increasing the swing counties to 9.  Meanwhile, he moved several blue counties further into the Democratic column; moving Osceola, Orange, and Miami-Dade into the 15-20% margin category.  He even managed to decrease the number of counties that favored the GOP by 20% or more.  Obama also made Duval, a large GOP county, much closer thanks to heavy minority registration and turnout drives with the African-American and student population there.

2008 Margin2008Swing

Obama’s 2008 win saw a major Democratic spike in urban counties.  Counties with modest or swing Dem margins (like Orange and Miami-Dade) jumped into heavily Democratic columns.  This was thanks not only to support from white suburbanites, but also massive registration increases in these counties, registering tens of thousands of minority voters.  Obama also notably weakened GOP margins in west coast suburban counties, but those counties, heavily white, would fall back into stronger GOP columns in four years. The shift in the urban counties was not just indicative of Obama’s large 2008 win, as the trend continued four years later.

Obama held onto Florida by just under a 1% margin.  While Palm Beach and Alachua fell out of the 20%+ Dem category, Osceola and Miami-Dade replaced them by becoming even more Democratic.  The number of counties 20%+ for the GOP increased and several red counties that fell within 5% in 2008 fell back to more solidly GOP territory.  The number of swing counties fell to a low of 5.   However, Obama’s improvement in major urban counties like Miami-Dade, while being able to stop major slippage in other urban areas like Hillsborough, Duval, and Orange, allowed him to win despite major fallout in the rural and suburban counties.

2012 Margin 2012Swing

Obama’s increases in Osceola and Miami-Dade were thanks to increased efforts with the Hispanic community and keeping African-American turnout up.  Heading into the election, huge registration efforts were done to shift the composition in these major counties.  These increases in minority registration helped counter backsliding with white suburbanites who has backed Obama in 2008. White liberals in the urban counties also held firm in the Democratic column.

Overall Trends

Looking at just the bar graphs shift from election to election, we see a massive growth in the heavily GOP counties.  Counties once modestly Republican have become entrenched for the party.  Meanwhile, few counties have actually moved to the Democratic column.  For Democrats, the key to success has been that the counties showing the largest shift in Democratic support have also been the largest counties.  In addition, the counties with the biggest Republican shift have been the least population.

The scatter-plot below shows how the margin has changed from 1992-2012 compared to the total votes cast in the 2012 Presidential election.  The counties which saw the Democratic margin improve (regardless of who won it) are in the blue end, while the counties who’s margin grew more GOP are in the red.


Every single county that cast more than 300,000 votes in the Presidential election has seen an increase in the Democratic margin.  Most of these counties are Democratic while others (Duval, Polk, Seminole, Hendry, Sarasota) are still GOP counties; just won by less than before.  Miami-Dade, the largest county, has seen a 20% improvement for the Democratic margin.  Many other heavily populated counties have seen major growth in the Democratic margins.   Meanwhile, the counties with the biggest growth in the GOP margin all fell well under the 100,000 votes mark, many casting less than 10,000 ballots.

Why the Big Counties got so Blue

Why have these big counties grown more Democratic?  Three key factors are in play.  1)  Increases of minority voters has been especially notable in the large counties.  2)  Urban whites maintaining Democratic support 3)  Democratic campaigns have set up major campaign operations in these counties.

Southeast Florida is the home of liberal whites.  Jewish retirees are a staple of the liberal white vote of Broward and Palm Beach.  Other counties like Pasco, Pinellas, and Leon have large shares of white voters who remain solidly Democratic. Meanwhile, the white suburbanites in other counties have trended more Republican while rural whites have taken a dramatic right-wing shift.  Not all white voters in urban counties remain Democratic, however.  Heavily white suburbs in Orange, Duval, and in the coastal areas of Broward and Palm Beach are much more Republican.  The issue of white voter shifts in Florida are tremendously complex and unique to each county (factoring in education, income, state of birth), but these points are some good generalizations.

The urban counties have seen major increases in the share of registered voters that belong to a minority group.  The scatterplot below shows how the non-white share of registered voters has increased compared to the total number registered.  The plot shows that while most counties have seen an increase, the largest counties have seen a higher-than-average growth in non-white registration.  Pre-2006, the state did not offer Hispanic as an option, leaving residents to chose white or other, making that data less consistent, which is why this focuses on 2006-Present.

Race Scatter

For Obama to win Florida in 2012, minority voters were crucial. While Obama won 42% of white voters in Florida in 2008, he only got 37% in 2012.  Obama held his support with the black vote and managed to increase Hispanic support by 3 points, from 57% to 60%.  Obama’s win relied on that increase in Hispanic support and an increase in the non-white vote in Florida.  Between 2008 and 2012, the white share of the vote fell from 71% to 67%, the black vote remained 14%, and the Hispanic vote went from 13% to 17%.   Miami-Dade’s growing margin for Obama from 2008 to 2012 is thanks to Obama’s larger share of the Hispanic vote and the fact that the non-white share of the vote fell 2%.  In Orange, despite Obama losing some ground with white suburbanites, the President held ground thanks to an increase in the non-white share of the electorate.  This didn’t work in all counties though.  Palm Beach saw it’s white share fall from 68% to 65%, but the President still did worse than in 2008.  In this case, Obama had a much larger drop in white support from the Northern side of the county, and the white vote was much larger in Palm Beach than other large counties.  The same issue applies to Pinellas.  However, the drops would have been worse if the non-white share of the vote hasn’t been increased from 2008.

Why the Rural Counties got so Red

The issue in the rural counties is a factor of national trends.  Florida’s rural counties, most located in the North, are cultural similar to the American South; which has seen an increasing right-wing shift in recent decades.  Bill Clinton was the last Democrat to play in much of the South.  Clinton’s strong southern backing was reflected in his wins and narrow losses in several rural counties. However, since Gore, the Democratic situation in the rural counties has only gotten worse as once ancestral-Democratic counties began to vote more GOP.  While Democrats used to maintain local control in these counties, that control is slipping as well.  In the rural counties near Lake Okeechobee, the Democratic platform on environmental issues has no doubt hurt the party in an area dominated by agriculture at the expense of the Everglades.  Hendry County is dominated by Agriculture but is notably less Republican than its neighbors.  Why is that?  The answer is that Hendry also has a larger non-white share of the vote than its neighbors.  Hendry’s white share of registration is 56%, well below the 70%-85% range its neighbors hold.  Racial dynamics is the one major factor that keeps rural counties like Madison, Jefferson, and Hendry from slipping heavily into the GOP camp.


These shifts in county support come from changes in voter allegiances, but also from changes in the electoral makeup of the state of Florida.  As urban counties grow more racially diverse and white liberals flock to urban centers, the big counties of Florida will continue to be steadfastly Democratic, offsetting loses in rural counties.  Only a few countries truly swing in Florida elections anymore at the Presidential level.  For both parties, the pathway to victory doesn’t lie so much in winning or losing which counties, but rather the margins they win or lose them by.

The Democratic Presidential Margin in All 67 Florida Counties


The Numbers Behind Ireland’s Historic Vote on Same-Sex Marriage

On May 22nd, Ireland became the first nation in the world to legalize same-sex marriage by popular referendum.  The vote for approval was 62%, amazingly high considering Ireland’s reputation as a more conservative, Catholic nation.  Polls constantly showed a major lead for the YES side of the referendum; which had the backing of all major political parties and prominent celebrities.

The measure passed in all but one constituency, the conservative region of Roscommon.  Meanwhile registering over 70% support in Dublin.  The strong showing, including wins in many rural/conservative constituencies in the North and Southwest, was a stunning testament to how support for same-sex marriage has skyrocketed in just the last decade.


The cartogram map below shows each of Ireland’s constituencies reshaped based on the total votes cast in the referendum.  Dublin, despite its small size, had 25% of the vote come from its constituencies.

Results Cartogram

The win for same-sex marriage was very impressive due to Ireland’s history as a more conservative, Catholic state.  So how did such a strong win become possible?

The Campaign

Once the referendum was scheduled, the YES campaign had a significantly larger amount of institutional support compared to the NO side.  All of the major political parties came out for the measure, including the Incumbent Prime Minister and the former President.  Both the Fine Gail and Labour Parties, who rule Ireland with a coalition government, as well as Sinn Feinn and Fianna Fail, who rule in the minority, supported the YES vote.  A poll commissioned by the Sunday times showed support for same-sex marriage by party allegiance.  All parties saw their members support the issue. The narrowest was for Fianna Fail, which was a conservative, center-right, populist-style party that appealed to rural Catholics.


The Press praised Fianna Fail for its support for the measure considering its base of support often came from areas less inclined to support the measure.  However, one Fianne Fail Senator just announced she was leaving the party for many reasons, one being she felt the party did not campaign aggressively enough for the YES side. According to Senator Averil Power, many members of the party didn’t back the measure, and those who did affirm the individual support in private refused to campaign for fear of losing rural Catholic votes in the next election.  In 2011, most of Fianna Fail’s best constituencies were the Northern areas, which backed same-sex marriage by the narrowest of margins.  Fine Gail was also lukewarm on support at first, with Labour really pushing the referendum.  However, once the measure proved to be popular, Fine Gail’s enthusiasm for the YES campaign grew.

The business community was also heavily in favor the measure, seeing it as a way to attract talent and other industries to the country.  Major international companies and local corporations endorsed the measure.  The major opposition group was the Catholic Church, but their opposition was much more tepid.

The YES campaigns worked hard to evoke emotion in the debate, running powerful ads and using personal stories to pull at the heart-strings of the voters.  The NO campaigners tried to argue the family would be damaged by the measure.  However, their arguments never resonated well with voters, especially in an era were other countries could be sited as legalizing same-sex marriage and showing no ill-effect.  The YES campaign ran one of the best political commercials I personally have seen for such a measure, running a family-friendly ad that appealed to all segments of the population.

The YES campaign worked to reassure voters of all stripes that voting for same-sex marriage did not run contrary to the values they held.  Whether it was the campaign or pre-held beliefs, there was not a great deal of class-divide of the issue. A poll commissioned by the Sunday Independent showed support for the measure across all class divisions expect rural farmers.


The poll’s nation-wide results were 53% Yes, 24% No, and 23% undecided (for reference).  While upper and middle class voters backed the measure more, the issue still had support with working class and manual labor voters.  Considering the strong win, its likely farmer’s did indeed reject the measure, but not overwhelmingly.

While the YES campaign had a great message and a well-energized team that canvassed neighborhoods and registered voters.  The NO side was overall much more quiet.  Some NO members were obnoxious, stereotypical fear-mongers.  However, many of the NO campaign leaders were generally respectful, they just didn’t have a strong message to reject the measure.  When the YES victory was clear, one of the major heads of the NO campaign tweeted out congratulations to the YES campaign.  Many of the hostile feelings during these referendums that have taken place in the US did not occur in Ireland, at least not to the same degree.

The Youth Vote

As is the case in America, the most supportive group toward same-sex marriage are young voters.  A March pollcommissioned by the Irish Times, showed voters 18-24 registering 84% for the YES side of the referendum.  For the YES campaign, ensuring young voters turned out would be critical in case the polling began to shift over the next two months.


The campaign worked to register over 60,000 youth voters in the last few weeks of the campaign, a monumental registration effort.  In addition thousands of Irish citizens, many of them young, came #hometovote to cast a ballot in the referendum.  Ireland does not allow vote by mail for those oversees, and the stories of people making the trip home became a sensation in Twitter and in the Press on the day of the referendum.

Polls leading up to election day showed that the support for same-sex marriage continued to be strong with youth and middle-aged and older voters, and narrowly trailing with the oldest sub-set of voters.  In the end, if there had been issues with youth turnout, the measure still would have passed.  However, the campaign made the right call in working to activate young voters as a way to ensure the measure had the breathing room if the polls narrowed.  The success in activating youth voters, increasing registration and turnout, was a major accomplishment for YES campaign and something the political parties of Ireland will not doubt aim to take advantage of.


Turnout was very strong for the referendum, topping 60%, making it the highest turnout in the last 20 years.  Turnout was very high in the Dublin region, an area critical for the YES campaign.

Results Cartogram Turnout

The margin was as strong as it was thanks to areas so supportive of the measure giving it such large support.  In contrast, some of the weakest areas for the YES campaign also had lower turnout.  The correlation is not especially strong, but it is noticeable.  Areas more supportive of the measure also happened to have higher turnout.


Again, the correlation isn’t very strong, many areas show high turnout with more modest support or show strong support with modest turnout.  However, there is a clear trend-line linking support and turnout to some degree.

The Church’s Tepid Opposition and Decline in Influence

The biggest contradiction to Ireland’s support for same-sex marriage was the fact that Ireland is so overwhelmingly Catholic.  As of the 2011 Census, 84% of Ireland identifies as Catholic, with 90% identifying as Christian.


In the aftermath of the referendum, many who followed the campaign have noted that while the Catholic Church took a position against the YES campaign, passing out a message for the Priests to read the Sunday before the election, the opposition was much more tepid than in the past.  The Church took a much stronger role campaigning against the 1995 referendum to legalize divorce, causing it to pass by only 0.6% despite leading polls by a large margin.  The map on the divorce referendum is below.  The map is from Irish Political Maps.  I highly recommend you check it out, the site has a great deal of maps and history of elections in Ireland going back as far as its time as British colony.

1995 -Divorce II

As the map shows, most constituencies rejected allowing divorce, with the measure only passing thanks to major support in Dublin and tepid support in other cities.  Many areas voted heavily against it.  Let me put up the same-sex marriage map again just to contrast how much has changed.  These were both issues rooted in religion and dealing with marriage; both with Church opposition.  However, the maps are radically different.  Areas that heavily rejected the divorce referendum supported same-sex marriage, even if with just modest approval.  It is a remarkable turnaround.


During this referendum, the Church’s opposition was fairly muted, not coming out till just before the vote.  The Church stated its opposition but the Church Leaders in Ireland did not decree/urge voters to cast ballots for the NO camp, leaving the decision to the individual.  The Archbishop of Dublin, an opponent of a YES vote, made headlines for admonishing those who used a harsh tone toward LGBT people.  Many Church leaders were noted for demanding civility and decrying homophobia as un-Christian.  While the Church sought to keep marriage as it was, it went out of its way to say it did not want to deny LGBT people the rights the desired.  Like the 1995 referendum on divorce, the Church did not state that supporting the position contrary to the Church was considered a sinful act.  The tone during the campaign further highlighted a growing evolution on gay rights within the Catholic Church, especially in the wake of Pope Francis. Many other prominent religious leaders in Ireland, including those who advocate for the homeless, backed  the YES campaign.  As the Washington Post said, these local leaders had more moral authority in the eyes of the voters than the Catholic Church hierarchy.

Another factor that cannot be overlooked is that the Catholic Church’s authority has been greatly weakened in Ireland.  The Priest Sex Abuse scandal has had a major impact in Ireland and has soured the public’s image of the Church, weakening its moral authority.  Weekly Church attendance is also at a much lower point than it was over past decades.  The line graph below shows the percent of the population attending Church weekly.  There is less data for Roman Catholic specifically.


Despite a bump in 2007 (which may be an outlier), the percent of those attending Church weekly has declines greatly over the last 25 years.  I pointed out the divorce referendum on the timeline to show how much higher Church attendance was then compared to now.

The decline in Church attendance does not mean Ireland is becoming less religious necessarily.  Most of the population still identifies with the Church.  However, it does mark a shift from a society that used to revolve around the Church.  The YES side could not have won without religious voters.  Indeed, according to one of the last polls released (by the Irish times), 45% of voters who planned to cast a YES vote considered themselves religious.  Of course what “religious” means is open to debate, but the fact is those voting YES did not consider themselves anti-religious or ambivalent.  Irish Catholics are clearly moving in a more liberal direction, something reflected in America, where 57% of Catholics support same-sex marriage as of 2014.  The Church’s tepid opposition and weakened authority no doubt made it easier for devout Catholics who wanted to support the YES campaign to do so.

Was There a ‘Shy No’ Vote?

One issue that came up in the last few weeks of the campaign was a concern that the polls, as few as there where, would be wrong.  The issue was reflected in past polling on referendums proving to be inaccurate, especially when on sensitive issues.  Notably, Ireland’s 1995 vote to legalize divorce passed with 0.6% of the vote after holding massive leads in the polls.  A more recent poll to abolish the Senate failed despite holding a comfortable lead.  The press began to pick up on the notion of whether or not their was a ‘shy no’ vote, people unwilling to tell pollsters (some of these polls done in person) they would vote no.  The issue of a ‘shy no’ voter came from the fact that the YES campaign used a very emotional and effective narrative to persuade people to vote yes.  The concern was that opponents of the measure would have feared being labeled intolerant or homophobic, and thus kept their opposition to themselves. When the vote came in with 62% support, it seemed the concerns of a ‘shy no’ vote were unfounded.  However, looking closer at the polling, it appears their may have been a ‘shy no’ vote after all.

The graph below shows the polls taken over the last two month.  The dots show the poll results, the line represents the trend.  In included two results the pollsters gave:  one where they asked voters YES, NO, or UNDECIDED, and the other where they gave results while excluding the undecided.  The decision to include the results that didn’t count undecided was made because there is always a possibility that many undecided voters simply did not go and cast ballots.  Studies have shown many undecided voters stay home if they cannot make up their minds on candidates/issues.  These studies are American-voter focused, so it cannot be assumed the same applies in Ireland.

In the last days before the campaign, four polls came out, three on the 17th.  The YES campaign maintained its lead, but things appeared closer than in past weeks.  When undecideds were counted, the YES campaign fell below the 62% they eventually won (represented by the straight green line).  When undecideds were taken out, YES support remained high 60s and low 70s.  Meanwhile, the NO side never really topped 30%, staying well below the 38% they got (represented by the straight red line).


If we were to say there was no ‘shy no’ vote, then how do we get to 38% opposition.  Using the last four polls, all of which came as the heels of the vote, I tried to find a way.  I took the NO percent for from each poll and added in 80% of the undecided vote to the NO camp.  Studies in the US have show undecideds in referendums, if they vote, tend to break for the status-quo.  80% is probably more generous than reality, however.


Even when using 80% of undecideds voting NO, two of the polls still don’t have a high enough NO vote.  The May 16th poll comes in just right, while the May 17th poll that showed an unusually high number of undecideds wouldn’t need it as high as 80%.  Could it be that the May 16th poll or the May 17th with high undecideds be right?  It is possible.  However, this assumes all undecideds would actually vote and none would stay home. The May 17th poll could be right if undecideds gave 60% of the vote to the NO camp.  However, I do find it suspicious that so many would be undecided so close to the vote, which no one else saw.  In addition, that pollster showed 29% of those 18-24 voters undecided, while every other pollster showed major support and little undecided among that group.

In the end, without a better understanding of Irish pollsters and their rankings in accuracy, it is hard to say for sure how much of a shy no vote their was.  I am inclined to believe, based on the data, that undecideds broke against the referendum in the end. This is based off past studies showing undecided vote status-quo. Cross-tabs from the pollsters generally showed rural, farmers, and older folks made up a larger share of the undecided vote, groups more likely to vote NO than YES.  However, the undecideds would have been forced to break overwhelmingly no, and this assumes others just didn’t stay home.  I believe their had to have been a minor ‘shy no’ effect to make up the difference.  However, this vote was clearly minimal, and didn’t effect the outcome similar to the Divorce referendum from 1995.


The Same-Sex marriage referendum in Ireland marks a major achievement for the equality movement worldwide.  Ireland’s vote broke stereotypes and provided a justification for other countries and their political parties to move forward on equality.  Northern Ireland and several Western European countries that do not have marriage equality will now feel pressure themselves to move forward.  The fighters for equality in those countries will have an example to look to as they wage their own campaigns of the coming years.


The Facts about the Jacksonville Runoff

The night of the March primary in Jacksonville, I wrote that Alvin Brown had a narrow path to victory.  The mayor needed to win over supporters of Bill Bishop, the moderate Republican who came in third place, and he needed to dramatically increase Democratic turnout.  When all was said and done on runoff night Alvin Brown narrowly lost re-election with 48.7% of the vote.  So what happened?

Bishop Endorsed Brown….kinda

Brown needed Bishop voters to back him.  However, once Bishop lost the primary, he opted not to endorse anyone at first.  Bishop had many socially-liberal backers that Brown needed to pull off a win, and an endorsement was key to that.  Bishop did finally endorse Brown, weeks later, after 24,000 absentee ballots had been returned. a poll days before the election showed Brown getting 42% of Bishop’s supporters while Curry got 46%.  If Brown wanted to win with 50% in the first round, he would have needed 44% of Bishop’s supporters.  With Bishop backers making a smaller share of the electorate in the runoff (due to increased turnout), Brown would either have needed to win Bishop backers by a larger margin or make the runoff electorate much more Democratic than the March primary electorate.  This leads me to point two.

Democratic Turnout Wasn’t High Enough

Over the last week, I was monitoring turnout very closely. Brown had won his election in 2011 thanks to Democrats outpacing Republicans by 10,000+.  However, in 2015 the Democratic turnout only outpaced the GOP by just over 3,000.  The data below shows how Democrats and Republicans did turnout-wise pre-election day, election day, and total.


Democrats saw the raw vote cast in 2011 fall while the GOP and NPAs saw increases. While Democrats still made up a larger share of the vote, it was not close to what they needed to win.   Both the share and raw vote margins were well below what Brown needed them to be.  As the 2011 data shows, much of the Democratic advantage came from election day while early/absentee was essentially a wash.  However, 2015 saw Democrats do much better in early/mail, while election day was worse.  Person-by-person turnout data (which I dont have yet) would likely confirm that many Democrats who voted on election day in 2011 voted early/mail this time, hence the Democrats doing so much better there.

For the last week, I monitored the shifts in partisan turnout daily to see how it shifted. Democrats started off in a hole due to the GOP’s strength with absentee ballots for Duval County.  Once early voting began, Democrats narrowed the gap and eventually took the lead.  By the end of early voting Sunday, they were at 5,500+ ballots over the Republicans.  Election day was critical for Democrats, they still needed a strong showing; which they did not get.

Democrats needed to be 48% of the vote, maybe a bit less depending on how NPAs voted.  However, while their share got to 47% by the end of early voting, it fell with election day’s numbers.


When examining the raw vote gap on a daily basis, the numbers showed Democrats making strong gains in the last days of early voting.  They stood at a 5,548 margin over the GOP by the end of Monday.  Considering Democrats netted 10K over the GOP in 2011 on eday, getting an extra 5K this time was entirely plausible.  However, the Democratic margin actually fell on election day.


There is one notable data point that has been overlooked.  Democrat’s share of the vote did fall from 2011.  However, so did the Democratic registration (as a percent) in the county.  NPA voters have surged from 14% to 22% between 2011 and 2015.  This meant that for Democrats to get 47% or 48% (optimal since NPAs make up a much smaller share of the votes cast) in terms of share of the vote, they would have had to dramatically outperformed their registration numbers.  In 2011, Democrats made up 47.4% of the vote despite only making up 43.9% of the registration. In 2015, Democrats made up 45.1% of the vote while making up 41.3% of the registration.


Democrats outperformed their registration share by 3.5% in 2011.  In 2015, they outperformed by 3.8%, so actually higher than before.  However, as the numbers/graph show, Democratic registration fell in 2015.

As for the GOP, they were 37% of registration in 2011 and did 5% better (making up 42% of the vote).  This year, they were 36% of the vote, and did 7% better (getting to 43% of the vote). So the GOP did even better than the the Democrats at increasing their share compared to registration shifts.

Who is to blame for the loss?

It was frustrating to see the anti-establishment folks instantly jump on the fact that Brown lost.  The Florida Squeeze, known for its attacks on the state Democratic Party, already has an article out slamming the party for the loss.  The article is wrong for a number of reasons.  The most important item is this.  It treated the election like Duval is a swing county and that the race was “toss of the coin” type race.  Well guess what, its not, and everyone knows it.  Duval is a moderate Republican county.  Obama aimed to win it in 2008 and 2012, coming up short both times.  The only major Democrat to win it (while running statewide) is Bill Nelson.  In fact, here is how Brown’s loss tonight stacks up against other major Democratic names.


Brown’s loss actually ranks better than Obama both times, Charlie Crist, and Alex Sink.  He only does worse than himself and Nelson.  Obama poured resources into the county but still came behind Brown.  Brown, meanwhile, was outspent by Lenny Curry by a large margin, anywhere from 3-1 to 2-1.

Does this mean Brown doesn’t share some blame for his loss? No he definitely does.  Brown’s acrimonious relationship with liberals, especially the LGBT community, did him no favors.  Brown also avoided Obama and Crist in 2012 and 2014, something that did not sit well with many.  Most importantly, Brown’s campaign was not as strong as it was in 2011 (you wont find many who don’t agree with that statement).  Fault lies with the campaign itself; not the State Party or any other Democratic group.  Plenty of blame can lie with the Brown campaign for not working harder to get Democratic turnout up and for allowing the coordination of it’s efforts with other groups/campaigns.

Individual Campaigns Matter

It should be noted that the night was not an overall bad night for Democrats.  Tommy Hazouri, the Democrat running for one of the At-Large City Council districts, won with 55% of the vote.  The Hazouri campaign had an all-star campaign staff and amazing consultants: including Eric Conrad, Jenny Busby, Karl Bash, Jessica Osborn, and Kevin Cate.  These talented folks ensured a Democratic victory county-wide despite the red tilt.  Individual campaign’s do…. in fact…. matter.

Equality Florida also backed Hazouri and several other Democrats and Republicans who backed a Human Right’s Ordinance.  In fact, with the results all in, HRO backers now have a majority on the city council.  Brown’s general opposition to an HRO resulted him not receiving Equality Florida backing/help.  Brown’s HRO opposition also hurt him with the business community, which wanted the ordinance passed to improve the county’s image.

Tommy Hazouri’s win compared to Brown’s loss can be seen in these two maps. Mayor Council 3 Part 2

Hazouri won all the precinct Brown did and won several others as well.  Meanwhile, Brown didn’t win any precincts that Hazouri’s opponent, Geoff Youngblood, won.

Mayor Council 3

Hazouri notably did better than Brown in Riverside and San Marco, liberal/arty regions of the county.  Hazouri also did better in the beach communities and in the Mandarin/upper class suburbs region.  Meanwhile, Brown did slightly better in some African-American regions, but only by a 2-3% margin.  Some of the Republican suburbs that bleed over from Clay County (the Orange Park region) also gave little more vote to Hazouri than Brown.  Considering Clay’s lack of elasticity, its not too surprising the margins weren’t much different between Brown and Curry there.  However, Hazouri did win several neighborhoods in that general area that Brown lost (areas less influenced by Clay).

Hazouri’s wins and where he did better reflect the strong campaign he ran.  I considered it a much better run campaign than Brown’s and the results bare fruit to that belief.

The nature of campaigns also brings up another point to remember about Alvin Brown. His win in 2011 was considered a major upset and his opponent, Mike Hogan, was widely considered to have run a bad campaign.  Hogan made many enemies with the business community with his opposition to investing in downtown Jacksonville, causing him to lose endorsements and money.  Hogan was also dragged down by Governor Scott, who pulled less popular than Obama at the time of the election.  Watching the race in 2011, I knew Brown had run a good campaign, but Hogan had also ran a bad one.  Curry’s campaign, imperfect for sure, was not as bad as Hogan’s.

I expect to see blogs jump up and down on the state party with no factual basis for the criticism.  We can all expect to see news articles on if this race has any national implications.  Honestly I don’t see it.  This was a bad campaign losing a race, with Brown still doing better than many past Democratic candidates.  If Hillary Clinton pulls similar numbers to Brown, then she will be winning the state of Florida.  The GOP should remember that.

Why the Wisconsin Supreme Court Election/Referendum Gave Conflicting Results

Wisconsin’s Supreme Court elections have been drawing a greater deal of scrutiny in recent years.  The state has been racked by partisan fights as Governor Walker and his GOP legislature have pushed the state further to the right.  The state Supreme Court has thus drawn more attention as it has become a decider on the constitutionality of several actions and laws pushed by the state.  Wisconsin elects its judges to 10 year terms, and currently conservatives hold a 4-3 majority.  The 2011 election, which nearly saw a conservative justice almost lose to a liberal challenger, drew a great deal of attention due to the narrow results and controversies regarding the reporting of those results.  Fast forward to 2015 and one of the liberal justices was up for re-election.  Justice Ann Walsh Bradley faced off against James Daley.  While the race was heated at times, it drew much less attention than the 2011 election.  Bradley led in fundraising, raising $380,000 to Daley’s $140,000.   Bradley won re-election with 58% of the vote; keeping the Supreme Court composition the same.

2015 Supreme Court

All the counties Bradley lost voted for Romney in 2012.  Meanwhile, the Justice won many Romney/Walker counties.  Bradley’s share of the vote was modestly tied to Obama’s, however, she did better than the President in nearly every county.


In addition the Supreme Court race, there was a referendum regarding the selection of the Chief Justice for the Court.  Before the referendum, the oldest member of the court, in the case a member of the liberal bloc, was designated the Chief Justice.  The referendum aimed to change the method of selection so that the Chief would be chosen by the justices themselves for two year terms.  In practice, this would ensure a conservative Chief Justice.  The measure was put on the ballot by the legislature with the backing of the Republicans and opposition of the Democrats.  The referendum was effectively a partisan fight, with many elected officials taking party-line stances for or against it.  However, the money race was largely one-sided, with the YES camp raising $600,000 and the NO camp raising just $80,000.  The referendum narrowly passed.

2015 Supreme Court Referendum

This meant split results for liberals in the state.  On the one hand, they held on to one of the Supreme Court seats.  However, they effectively lost control of the Chief Justice position.  Any effort for a Pro-Bradley/Anti-Referendum slate were clearly underfunded or poorly organized.  However, there is some trend in the voting.   When analyzing the county by county results, the support for Bradley and opposition to the referendum were modestly tied together.  As support for Bradley grew, so did opposition to the measure.  However, almost no counties fell on or near the red-line in the scatter-plot below; which would have indicated equal support for Bradley and opposition to the referendum.


The movement to reject the referendum did 11% worse than Bradley did.  Had the effort only done 8% worse, it would have succeeded in killing the measure.  The green shade in the above scatter-plot represents the 8% gap that the counties could have fallen in to ensure success.  No county needed to reject the measure more than they supported Bradley (putting them under/to the right of the red line), a sufficient number just needed to fall within the green by shifting further to the right on the plot (while staying at the same level on the y-axis).

The map below shows where the gaps between Bradley and a NO vote were strongest.   Some of the most Republican areas had the smallest gap.

2015 Supreme Court Difference

There are lose ties between the gap support for Bradley.  The trend is not especially strong, but it is there.  Counties heavily in favor or Bradley saw a larger gap between her support and a NO on the referendum.  As the map above and plot below show, no county gave NO a higher percent than it gave Bradley.


The most likely culprit is not a split in the Democrat/liberal vote, but rather that swing voters went more toward supporting the measure.  As mentioned, Bradley won several areas that are more Republican in traditional elections.  When I checked to see if the gap between Bradley and NO increased based on Obama’s support in 2012, I found no clear trend.  It was not the case that traditional Democratic counties were more accommodating the the measure, it was that many of Bradley’s strong counties were so strong because she was winning over GOP voters; not because the counties were so Democratic.  Dane County, home to Madison, was Bradley’s strongest county.  The county is always heavily Democratic, and also featured one of the smallest gaps between NO and Bradley.  The gap was higher in areas were Bradley outperformed the President.

When I compared the gap between Bradley’s support and NO on the referendum to how Bradley did compared to Obama, I found that as Bradley improved over Obama more, the more the NO vote under-performed Bradley.



So what appears clear is that many voters who would not support Obama were willing to vote for Bradley, but also for a referendum shifting the court’s leader to a conservative.  In fact, while there is a strong correlation between Obama and Bradley’s vote share, the correlation between Obama and a NO vote for the referendum were slightly stronger.  The scatter-plot below shows how the NO vote and the Bradley vote tie to Obama’s vote (the red line being a perfect match).  The NO vote runs under Obama for the most part. Statistically, the NO vote had a greater tie to Obama than Bradley, with a coefficient of 0.63 compared to the 0.59 for Obama/Bradley (1.0 being a perfect correlation).


From the data, it appears clear that some of the voters willing to vote for Bradley, perhaps due to issues were her opponent or support for the Justice, were also more GOP-leaning and thus willing to back the referendum.

However, it should be noted that some of the Pro-Bradley, pro-referendum voters were probably also moderates who felt the referendum sounded good.  The referendum backers had much more money and framed the issue as a more Democratic method.  Many supporters of the measure likely did not know the partisan intent.  Another factor in the successful package is likely that GOP turnout was higher than Democrats.

The results of the Wisconsin Supreme Court election and referendum do give conflicting results.  However, both races still saw clear partisan trends.  Bradley won by so much thanks to the support of voters who don’t normally vote Democratic.  However, those same swing voters put the referendum over the top, ensuring a conservative Chief Justice on the State Supreme Court.