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.

Quick Thoughts on Jacksonville’s Upcoming Runoff

Jacksonville had its first round of voting Tuesday.  Democrat Alvin Brown, first elected in an upset win in 2011, is seeking re-election in a hard-fought contest.  Brown’s main challenger is Republican Lenny Curry, the former Chairman of the Republican Party of Florida.  The race has been heated so far.  Brown is facing stiff headwind due to the county’s Republican Lean.  However, the Mayor’s approval rating remains above 50% in recent polls.  With four candidates in the race, it was widely believed that Brown and Curry would advance to a runoff election.

As expected, Alvin Brown and Lenny Curry advanced to a May runoff.  Brown secured 42.6%, with Curry a few points back at 38.4%.  Another Republican, Councilman Bill Bishop, secured 16.8%.

Brown’s biggest base of support was the African-American community.  However, he did well in white suburbs as well.  Curry did especially well in the outlier areas and traditional GOP bases of support.  Bishop did best in his council districts, located on the northeastern edge of the county, and did well in many central duval precincts outside his district.


While the GOP vote passed 50%, it is incorrect to look at this race in a purely partisan lens.  In two of the five at-large council races, Democratic candidates totaled more than 50%.  Each race has its own dynamic based on the candidates and issues.  In the Mayoral race, Bill Bishop secured a good deal of support from non-Republican sources.  Bishop’s social liberalism and fiscal conservationism earned him a good deal of support from NPA voters. Bishop’s support for the proposed Human Rights Ordinance of Jacksonville also resulted in several LGBT groups endorsing him for Mayor. The general anecdotal story from the area was that Bishop was winning many socially liberal voters.

The maps show that Curry narrowly won the Riverside/San Marco areas (south/middle part of the county).  However, he won it with less than 50% and Bishop did better in this region than most areas.  The region has many white liberals who backed Bishop for his social liberal views.

Curry won the beaches, the upper class suburbs (Orange Park, Mandarin) and the west side, which is much more rural.

Public polling has been scarce, however, a February UNF poll gives a little insight into where Bishop might have been receiving votes.   In the poll, Bishop was the choice of 8% of Democrats, 14% of Republicans, and 21% of NPA voters. Among voters who felt they could identify Bishop’s ideology, 6% ranked him liberal, 10% ranked him moderate, and 5% ranked him conservative:  most not knowing what they felt his ideology was.  This is one poll and is outdated.  However, from it, it’s not hard to see Bishop taking in moderate voters, liberals, and conservatives.

This poll is backed up by precinct data from the primary.  I examined where Bishop did best and found as he did better, Brown under-performed Charlie Crist more.  Brown’s 42.6% outpaced Crist’s 41.5%.  However, many precincts saw Brown dramatically under-perform Crist; all the while outperforming Crist in other regions.  The biggest factor in Brown under-performing Crist was Bishop’s percent of the vote growing.

I looked at this factor on a scatter-plot.  I focused only on largely white precincts to get a better idea of the effect; leaving African-American precincts out of the scatter-plot.  African-Americans overwhelmingly backed Brown, giving Bishop and Curry next to nothing.  My goal was to see who Bishop was taking the most votes from among white voters.


The scatter-plot shows a clear trend.  As Bishop’s percent increased, Brown under-performed Crist more, while Bishop’s weaker precincts showed Brown beating out the former Governor.  The coefficient came in at 0.76.   The coefficient tell us how strong the correlation (ties) between two data points are, ranging from 0 to 1 (1 being a perfect correlation).   The result of 0.76 means the correlation is strong.

I also examined how Bishop’s strength would effect Curry as he related to Rick Scott’s percentages.  After all, a third person doing well would hurt both candidates as they related to the Gubernatorial results from several months back.  The results show that, indeed, Curry does worse than Scott as Bishop’s vote increases.  However, the correlation is much weaker.


When examining Curry’s under-performance, more precincts are scattered around, rather than falling on a clear pattern.  The coefficient is also much weaker, 0.26.

From both scatter-plots, it’s clear that Bishop took from both sides.  While it is tempting to say he took more from Brown than Curry, that can’t be known for sure.  The same UNF poll from February said that the second choice of Bishop supporters was Curry by a 48% to 31% margin.  This may have changed over the last month, especially since Bishop ended up with more than the 11% he was polling at.

For Brown, one thing is clear, he needs to get the support of as many Bishop voters as possible.  Brown would have needed 44% of Bishop’s votes to avoid the runoff (granted this does not factor in how the NPA candidate’s votes would have gone).  Heading into a runoff, Brown just needs to make sure the Bishop voters who are more aligned with him show up.  He cannot count on the Bishop voters more aligned with Curry to stay home.  Curry’s extreme conservationism should be utilized by Brown to ensure Bishop’s liberal supporters return for the runoff.  Brown definitely benefits from Curry’s unabashed conservatism, the same thing that derailed Brown’s 2011 opponent.

One final factor in how the runoff goes will be the turnout.  Democrats did a good job getting the vote out for this first round.  Over 3,300 more Democrats than Republicans showed up for this March primary.  This was a strong improvement from the 2011 primary where the Republicans outpaced Democrats by just over 1,000 people.

However, when Brown won the runoff in 2011, Democrat’s turned out more than Republicans by a margin of over 10,000.


Democrats made up around 44% of the vote in the 2011 primary, but made up to 47% in the runoff.  In this first round of voting, they reached 45% of the electorate.  For Brown, he needs to replicate that huge runoff surge as much as possible.

For Brown, there is a clear pathway for a runoff victory.  However, it will not be any easier than winning the 2011 runoff was.  A turnout push and securing Bishop backers are key to a victory in May.

Maps and Quick Thoughts on Broward’s Municipal Election Results

As a Broward County resident for the first 19 years of my life, Broward County and its politics are still very close to me.  The county, second largest in the state, has 35 cities and towns.  Many of these cities host elections in the fall of even-numbered years to coincide with major races.   However, many cities continue to hold elections in the spring of odd and even numbered years.  March 10th saw the latest round of municipal elections for the county; with eight different cities going to the polls.  The following article will contain a map of each race’s result and a quick summary of the results and events surrounding the election.  While some races were fairly quiet, others were major battles for the future of the cities.

Plantation Mayoral Election

Plantation’s mayoral election was the main event in Broward on election night.  This suburban city leans Democratic but is a swing city compared to many of its bluer neighbors.  Republicans had success in the city in 2013, with one Republican and one Independent winning seats on anti-tax platforms following a revolt over a large increase in property taxes.  The city has a strong mayor form of government and Republicans wanted to take control of the Mayorship in order to get stronger control.  Commissioner Jerry Fagden, who had been elected to the commission with 60% in 2013 as part of the revolt, challenged Democratic Mayor Diane Bendekovic.  Bendekovic and Fadgen squared off in 2011 for the seat as well.  After a closely fought battle in which the local Republican and Democratic Parties got involved, Bendekovic held on with a narrow in.  Fagden narrowly won election day voters but lost on the absentee votes.


Bendekovic ran up strong margins with the African-American precincts in the east and many western and central suburbs.  Many communities were split, with all precincts giving the mayor at least 40% but no more than 70%.  The narrow win was a victory for local Democrats.


Plantation Council 2

With Fadgen running for Mayor, he had to resign his council seat.  This gave Democrats a chance to take the seat after failing to do so in 2013.  Democrats had a problem, however, where four Democrats filed along with only one Republican, Rico Petrocelli, a former commissioner.  This left Democrats worrying that a split vote would lead to a Republican win.  Petrocelli, however, had a great deal of personal and financial baggage.  Many Democrats rallied around lawyer Louis Reinstein.  Democrat Peter Tingom, the commissioner who lost to an independent in the 2013 elections, also filed to run.


Tingom managed to eak out a win, no doubt helped by his name recognition, while Petrocelli came in second.  Reinstein’s loss was a disappointed for many Democrats.  However, the split field made the results hard to predict and shape.  Claudette Hammond, and Jeff Holness, both African-American, split the eastern end of the city.  Holness, the cousin of County Commissioner and wanna-be kingmaker of Broward, Dale Holness, had garnered 34 of the district in his 2013 run against Fagden, but this time finished fourth.


Miramar Mayoral Election

The Miramar Mayoral Election was right behind the Plantation Mayoral in terms of major implications.  Longtime Mayor Lori Moseley was seeking re-election against an incumbent commissioner, Wayne Messam, and former commissioner, Alexandra Davis.  Moseley has served as mayor for the city since 1999.  Since that date, the cities’ population tripled from 40,000 to over 120,000.  The city has become much more diverse, with the largest growth being among African-Americans.  By 2010, Moseley, who’s white, has been representing a city that is 46% black, 37% hispanic, and 11% white.  Both Davis and Messam where African-American, and the cities’ last white commissioner, a Republican, was ousted in 2013.  The race was seen as close by most observes, with each candidate having a chance at winning.


Moseley ended up losing the election, coming in second to Messam.  Moseley did best in the western end of the city, which is predominantly white or Hispanic and did worse in the eastern, African-American side of town.  Davis only won a few precincts on the eastern end.  Moseley’s loss can be partially attributed to race and the desire of African-Americans to be represented in the Mayors office.  This no doubt moved some votes.  However, the cities population growth also likely helped aid Moseley’s loss, as newer residents had little reason to support a mayor they didn’t know.  In addition, issues over a city bond to fund new projects was a key debating point.  Moseley did not favor a large bond while her opponents did.  Messam praised Moseley as a good mayor but argued for a new direction as the city continues to grow.  With 16 years on the commission, many voters who liked Moseley may have thought it was simply time for a new vision for the city.  With the results for mayor and the city council elections, the entire commission will be African-American for the first time in the cities’ history.


Miramar Commission Districts 1 and 4

Two city council seats for Miramar were up for election.  District 1 had been vacated by Alexandra Davis.  Davis had initially left the seat to run for county commission in 2014.  Davis was badly beaten in the Democratic Primary by Incumbent Barbara Sharief that August.  Davis initially decided to run for her empty city council seat again.  However, Sharief’s husband, Maxwell Chambers, opted to run for the seat.  The move was no doubt influenced by a dislike for Davis.  Davis eventually decided to run for mayor, where she came for third.  Chambers was the front-runner for the crowded city council field, getting a good deal of support from local leaders and donors.  Chambers went on to win with 35% of the vote.


Chambers won by winning many African-American areas, splitting the African-American vote with Norm Hemming.  Alejandro Casas, the only Hispanic in the race, did well in the west and came in third.

In the district 4 race, left open by Messam running for Mayor, three candidates faced off for the seat.  All three had local ties to the area, but none had key regional support.  With all three candidates being African-American, the east-west split the city has seen in other elections did not manifest itself.  Darlene Riggs, a businesswoman, won the seat by a fairly comfortable margin.


Riggs won the seat by winning in African-American and Hispanic/white areas, winning all but four precincts.  She had the strongest win in Miramar that night.


Coconut Creek Commission District E

Only one commission seat was contested in 2015 in Coconut Creek.  The seat was opened up when Commissioner Lisa Aronson filed to run for the 2014 county commission election, which she lost to Democrat Mark Bogen (whom you’re truly worked for).  The race pitted Steve Harrison, a past candidate, against Joshua Rydell, a young newcomer to coconut creek politics.  Rydell racked up the backing of the political establishment, including former Representative Jim Waldman, current Representative Kristin Jacobs, and many members of the city council.

Coconut Creek

The results were a resounding win for Rydell.  He won all but one precinct.  Over half the vote came from the Wynmoor Condo Community (the second most-southern precinct on the east side of the city), which also gave Rydell over 80 of the vote.


Davie District 1

Davie only had one election on its ballot, with incumbent Mayor Judy Paul unopposed for another term.  Commissioner Bryan Caletka beat James Moore in a rematch from 2012.


Caletka held on to his seat with a comfortable margin, similar to his 2012 re-election.  Moore only one the eastern edge of the city.  Moore made traffic issues a key theme of his campaign.  The area includes University Blvd, an area known for heavy traffic issues. The road actually divided pro-Moore and pro-Caletka precincts in the north.  This is the last election where Davie will host elections in the spring.  Future elections will coincide with statewide races.


Deerfield Beach Districts 1 and 2

Deerfield Beach has been home to some contentious politics as of late.  Mayor Jean Robb has been feuding with the city council over several issues (quick summary, the council is right).  Robb won the mayoral election in 2013 by 27 votes.  Two council members were up for re-election, with Robb hoping both would lose.  Commissioner Joe Miller faced off against Ron Coddington.  Coddington was very flawed, having been removed from the Marine Advisory Board after he accused Miller of using his elected position to benefit his business (of which there is no truth or even suspicion).  Coddington’s campaign was filled with controversies and mis-steps and he handily lost to Miller.


The district 2 race saw a greater likely hood of the incumbent losing.  The seat lies in the African-American area of the city the white incumbent, Ben Preston, was facing off against Gloria Battle, who is African-American.  This was also the same region that had backed Robb heavily in her Mayoral Campaign.  Battle garnered controversy when residents in the retirement community, The Deerfield Beach Palms, complaining of over-zealous individuals trying to collect absentee ballots in favor of Gloria Battle.  Issues of absentee collection and even fraud permeated in the mayoral election and seemed to be rising up again here.  It could be argued the Mayoral election was won via questionable absentee ballots.  The issue is way too complex to discuss here.  However, Battle ended up winning by a large margin, enough that any fraud issue is unlikely to have effected the result.  Battle won in the same areas that backed Robb in 2013.


There has been some talk that Preston didn’t listen to his consultants and didn’t take the threat seriously enough.  Even with Preston’s loss, Robb is still outnumbered on the city council.


Lighthouse Point Districts 2 and 3

Lighthouse Point is a quiet bedroom community, and one of the few solidly Republican areas in the county, on the upper coast of the county.  The city of 10,000 has its own police force and likes to maintain its own closed in community where everyone knows each other   Both council elections saw former commissioners run for seats.   In district 2, Incumbent Commissioner Michael Long defeated former Commissioner Tom Hasis. Hasis initially lost his seat years back and had since run for and lost in another district race. Voters are clearly wary of Hasis, who has a contentious demeanor that conflicts with the quiet city.  He lost by a large margin.


In the district 3 race, which was open, former commissioner Susie Gordon, who had lost her seat in 2012, opted to run again.  However, the election went to 35 year old lawyer Jason Joffe.  Gordon came in third.


Not a good night for former commissioners.


Hillsboro Beach Commission

Hillsboro Beach is the definition of small town.  The city is a strip of land that goes along the coastal A1A road.  The town is home to many wealthy coastal homes and part of is often refereed to as Millionaires’ Mile.  The town is home to rich residents and is one of the few Republican areas in the county.  It is right next to Lighthouse Point and population is just over 1,800.  The city is just one precinct.  Voters go to the polls and everyone runs for the council, with the top three candidates getting spots on the commission.  Several incumbents and new candidates ran for the council.  The biggest issue for the town has been beach erosion, which has sparked much debate in terms of a viable solution.

Hillsborough Beach

In the results, Deb Tarrant, a newcomer who has argued the commission needs new leadership, came in first, commissioner Victoria Freaman came in second, and former Mayor Carmen McGarry came in third.  The current Mayor (commissioners select the Mayor) Claire Schubert, came in second to last.  Tarrant says her first goal is to see about getting a city manager appointed to handle the issues before the city, especially regarding beach erosion.  The idea is likely to spark debate and would require a referendum.  Even in small towns like this, the local politics is never quiet.


Ft Lauderdale District 2

Ft Lauderdale Commissioner Dean Trantalis had an easy re-election over local businessman, and political newcomer, David Tabb.  Trantalis was the first openly gay commissioner in Ft Lauderdale history when he was first elected in 2003.  Trantalis won his current seat in 2013, beating former commissioner Charlotte Rodstrom, after being off the commission for a few years.  In a quiet campaign, Trantalis was easily re-elected.


The February mayoral election for Ft Lauderdale drew more attention over the cities ban on feeding the homeless.  Even though Trantalis supported the measure, the issue did not appear to affect the narrative of the city council campaign compared to that of the mayoral election.



Several elections in Broward were sleepers and saw the status-quo maintained.  However, Miramar saw major changes in its city Government.  Democrats struck back in Plantation after an awful 2013.  The contentious politics of Deerfield will continue on, with both sides of the debate likely to clash in the Mayoral Election in two years.  As more cities move their elections to the fall to coincide with statewide races (Davie and Plantation will be doing so), there is a chance these local issues will get drowned out by the larger campaigns.  Until then, Broward always provides some interesting observations with its spring elections.

Florida Senate District 7: A Case in Compact Gerrymandering

Redistricting History

In Florida, a great deal of attention has been paid to the issue of gerrymandering during the 2012 redistricting process.  The state has been subjected to lawsuits over its Congressional and Legislative Maps by a coalition arguing the legislature violated new redistricting rules passed by voters in 2010.  A lawsuit in the summer of 2014 forced the legislature to alter several of its congressional districts.  The changes were small, but the lawsuit opened up a much larger can of worms.  Through the lawsuit, the consulting firm, Data Targeting, which has worked with the legislature and Florida GOP, was forced to turn over 500 pages in emails and documents detailing its involvement in the redistricting process of 2012.

In 2010, Florida voters approved constitutional amendments to bar partisanship from factoring into redistricting.  Districts were to be compact where possible, not designed to protect incumbents, drawn to protect communities of interest, and be drawn in a non-partisan manor.  The Florida GOP did not want these amendments passed.  However, once the results were in, they pledged to abide by the rules. However, when the 500 pages of sealed documents from Data Targeting were released, they showed the legislature had used firm to submit maps using average citizens in order to mask the firm’s involvement.  Email correspondence showed the firm working to make sure districts kept incumbents in strong districts and to ensure a GOP dominated map.  Emails showed the firm intended to keep discussions off email when possible to avoid a paper trail.

The Florida Supreme Court is scheduled to hear the case of the state legislative districts and whether the fair district amendments of 2010 were broken.  Meanwhile, the legislature, knowing they have been caught red handed, intend to argue that the fair district amendments were not legal; even though they are in the Florida Constitution.

The Issue of Compactness

One of the main focuses has been the “compactness” of the districts.  Florida has been known for winding districts that stretch for miles.  When the legislature passed its new redistricting maps, the Supreme Court actually invalidated eight of the forty senate districts for not adhering to the compactness rules. In Southeast Florida, the districts out of Palm Beach and Broward were questioned.  Specifically, district 29, which was supposed to be Ellyn Bogdanoff’s district, was thrown out.



The district was designed to save Bogdanoff’s re-election by creating a coastal district that took in GOP communities.  However, the State Supreme Court said the district violated compactness rules and a redraw was forced.  The final map forced Bogdanoff into a much more compact district, labeled number 34.



The district put Senator Bogdanoff against Democratic Senator Maria Sachs in a member-on-member fight for the seat.  With the district more Democratic due to taking in Democratic area’s further off the coast, Sachs won the 2012 election.  This as not what Republican leaders wanted.  However, the State Supreme Court had forced their hand

The Fair District Amendment’s Compactness requirement put a damper on the legislature’s redistricting strategy.  With the legislature’s ability to create odd districts curtailed, except in case of minority districts, many may have thought the resulting maps would be free of partisan intent.  However, the legislature, or should we say its consulting firms, managed to create a Republican-leaning map that featured districts much more compact than the map from the last decade.  There are several examples of compact districts that were still clearly designed with a partisan intent.  I like to call this process Compact Gerrymandering.  This article will focus on perhaps the most simple and effective version of compact gerrymandering in the state:  Senate District 7.

Senate District 7

Senate District 7 lies in North Central Florida and is made up of just three counties:  Alachua, Clay, and Bradford.  It is one of just two senate districts to not split any counties.


The district’s shape is reasonable and by keeping three counties contained in one district, lawmakers get to argue they are preserving communities by keeping entire counties from being carved up. After all, the voters of a county have a common interest and want that voting power as strong as possible as a unit. A county has limited power when its voter are split among several legislative districts.

However, Senate 7 is still a prominent example of partisan gerrymandering. The district takes Alachua, a Democratic county that is home to Gainesville and the University of Florida, and pairs it with Clay County, one of the most Republican counties in the entire state.  Alachua has slightly higher registration that Clay, however, Clay’s much more Republican that Alachua is Democratic.  Bradford rarely decides elections in the district due to its low population size.  The Party registration breakdown is below.

FL SD7:  Compact Gerrymandering

Clay County is made up of the Jacksonville suburbs and other high-end communities. The voters there never vote Democratic.  Despite being just over 50% Republican, the county rarely gives Democrats more than 30% of the vote; meaning the NPAs vote GOP much more than Democrat.   In 2006, Clay was one of only nine counties to vote against Bill Nelson in his landslide re-election. Clay’s voters have little in common with Alachua’s. Alachua is made up of students, African-Americans, white suburbs, and rural whites. Alachua would be better paired with Marion county or several of the rural counties around it. Clay fits better with Duval and St Johns. However, by linking the two counties together via small, rural Bradford, the legislature has created a district that is modestly, but stubbornly, Republican. Bill Nelson narrowly lost the district in his 2012 re-election (while winning the state by double-digits) thanks to Clay’s rigid Republicanism. This district shuts out the voters of Alachua, who can never override Clay’s partisan bent.

I examined several recent elections in the district to get a feel for just how strong the partisan bent is.  First, lets look at the 2012 Presidential Election.


While Obama won Alachua with 57.9% of the vote, he lost Clay with 26.7% and Bradford with 28.6%.  The dark red of Clay stands out on the map.

In that same election, there was an open state senate contest.  However Democrat William Mazzota lost to Republican Rob Bradley.  The map barely looks different from the Presidential contest.

3State Senate

Mazzota ran into the same problem that Obama did, a big brick wall in Clay county, where he only received 25.8% of the vote.  Obama and Mazzota ran fairly close to each-other in most parts of the state.  Mazzota managed to over-perform Obama mostly in Bradford.

Obama Mazzota Compar

Unfortunately for Mazzotta, the one place he did better than Obama in was also where the least vote comes from.

Jumping ahead to 2014, the Gubernatorial race told a similar story for the district.  Democrat Charlie Crist lost the district, doing worse than Obama thanks to Clay county.

1 Governor

Crist got 56.4% in Alachua, around 2 points worse than Obama.  Meanwhile he got 23.4% in Clay, 3 points worse than the President.  Crist did manage 30.4% in Bradford, the one county he improved over Obama in.  Crist also suffered from Clay being worth more of the vote than it was in 2012 thanks to turnout disparities in the midterms.  Clay was worth 42% of the vote in 2012, but went up to 44% in 2014.  With such bad margins in Clay, that slight uptick can make a big difference.

The worst Democratic performance in the district was the CFO race, where Democrat Will Rankin fell below 40% in the district.

0CFO 2014

.Rankin got 52.4% in Alachua, 25.5% in Bradford, and completely tanked with 20.5% in Clay.  Rankin’s results serve as a basement in the district for the last two cycles.

The only Democrat in recent years to come close to winning the district is Bill Nelson.  Nelson won re-election by 13 points in 2012.  However, he lost the district by 56 votes.

4US Senate

Nelson got 62% in Alachua, 39.5% in Bradford, but only got 33% in Clay.  Clay’s Republicanism held despite Nelson’s strong statewide win. Nelson managed to win a few precincts in the county, but still was crushed in much of the area.  Clay’s Republicanism kept Nelson down enough to just narrowly lose the district.

Back in 2006, Nelson won the district thanks to improved margins in Alachua and Bradford.  However, even in the 2006 landslide that saw Nelson get 60% statewide, the Senator only managed 38% in Clay.

In 2012, Nelson got close to winning because he over-performing Obama across the district.  Nelson’s gains were weakest in Alachua and strongest in Bradford.  Nelson did 4 points better in Alachua, 6 points better in Clay, and 10 points better in Bradford.

Obama Nelson

Alachua may be Democratic, but it is not elastic enough to give Democrats the large margins they need to win the district.  While recent Democrats peak at 62% in Alachua, they crater to 20% in Clay.  Clay is actually more elastic than Alachua, but in a bad way: giving Democrats margins from bad to terrible.  Nelson’s 06 win (and 38% in Clay) is a fluke caused by a disastrous Republican opponent.  Nelson’s 2012 percentage of 33% in Clay is the best the party can hope for, and with Alachua not willing to break higher than low 60s, winning the district is extremely unlikely.

Elasticity in the District

For a Democrat to win SD7, they would need to run ahead of the President.  Any winning Democrat would have to be a moderate that could win over rural voters and Republicans even while they vote Republican for other races.  Too see if enough of these voters exist in the district, I took at the look at the elasticity (the willingness to cross party lines) of the district.  Looking at the results for the five elections examined above, I examined each of the five Democrat’s precinct results and mapped the gap between the lowest and highest performing Democrat in each precinct.

Diff Gap

As the map shows, the precincts in Alachua often see the lowest gap between the best and worse performing Democrat.  Clay and Bradford have the biggest gaps.  In most of these instances, the gap is caused by Nelson doing much better than other Democrats.  However, Nelson is still badly losing most of these precincts.  The precincts that performed best for Democrats have some of the weakest elasticity.   The scatter-plot below shows precincts by their percentage for Obama and the gap between the worst and best performing Democrat.


As the plot shows, as precincts become less favorable toward Obama, they generally have a greater gap between Democratic candidates.  Now this can mean some Democrats managed to run ahead of the pack (like Nelson).  However, it also means other Democrats (like Rankin) fell off even more in dark red territory.  Meanwhile, strong Obama areas (60%+) did not give a strong Democrat like Nelson a much higher share of the vote.

The scatter-plot below shows how Nelson, the strongest performing Democrat, and Rankin, the weakest performing Democrat, compare to Obama’s percent of the vote.  The red line signifies a perfect correlation (same percent).


As the plot shows, Rankin runs worse than the President in most precincts while Nelson runs ahead.  However, the gap between Rankin and and Nelson is highest in the least Democratic areas.  In the heavily GOP region, Nelson is able to run higher than Obama while Rankin runs worse.  Rankin stays below the President as precincts become more supportive of the President while Nelson’s advantage narrows.  By the time we reach the precincts with the strongest Obama percent, Nelson is barely doing better than the President, all the while Rankin still trails.  So what does this mean?   It means that while a weak Democrat like Rankin can do worse than the President in a very dark red part of the district, a strong candidate like Nelson cannot do better than the President in heavily Democratic territory.  Therefor, there is limited room to improve but always room to do worse.  This is not unique to just District 7.  In general, some of the most Democratic areas have limited elasticity because they are heavily African-American, and already have voters casting straight-Dem ballots.  In some of the most Democratic areas, there are simply too few swing voters for someone like Nelson to get.  Therefore, a moderate candidate like Nelson would need to wrack up even stronger improvements in red territory; most likely moderate white suburbs or rural ancestral Democrats.  Bradford, an ancestral Democratic territory, has the right type of voters to carry this feat.  Bradford, like many rural counties in the state, is still Democratic at the local level and has voters willing to split tickets and vote for a Democrat they like; even while giving Democratic candidates for President very little support.  Despite losing Bradford, Nelson did 11% better than Obama did in 2012. However, the problem is that Bradford is too little of the vote, often around 5% of ballots cast.  Nelson only managed to do 6% better in Clay and 4% better in Alachua, were most of the vote takes place.  Clay is not going to provide the level of ticket-splitting that a moderate Democrat needs to win SD7, making a Democratic win there almost impossible.

Before redistricting, the Senate district based out of Alachua, then District 14, was a swing seat.  The district had narrowly voted for Obama in 2008 but voted GOP on other occasions.  The old district boundaries can be seen below.Old14th

The key to the district’s swing nature was that it paired reliably blue Alachua with several rural ancestral Democrats.  The voters of the surrounding counties are largely Democratic in registration and vote Democrat at the local level and largely Republican at the top of the ballot.  However, Obama could muster some votes out of these areas and candidates like Nelson could improve on Obama’s margin by a significant amount.  In 2012, Obama would have lost the district with 47.8%.  Nelson would receive 54.5% thanks to major improvements over Obama in the rural counties.  The key to that district being in play is the lack of Clay and the inclusion of the rural counties.

Clay was willing to give Nelson better marks than Obama, but only by a few points.  The county’s suburban Republican voters are brick wall for Democrats, even for moderates like Nelson.  There is no Democrat I can think of who can break that brick wall in Clay.

The Growing Problem of Clay County

Clay’s strong Republican lean makes any race in SD7 a fight between it and Alachua. The bar graph below shows the Democratic candidate margins for the three main elections in 2012 (President, US Senate, and State Senate).  Democrats rack up huge margins in Alachua, lose in small Bradford, and shed tens of thousands of votes in Clay.

Election Bar graph

For Obama and Mazzota, the problem was was they lost Clay by more votes than they won Alachua, despite the fact Alachua cast more votes than Clay.  Nelson’s race was the one instance were Bradford decided the election.  Nelson actually managed to win 2,000 more votes in Alachua than he lost in clay.  However, by losing by just over 2,000 votes in Bradford, he narrowly lost the district.  Nelson got close to winning by severely cutting the number of raw votes lost out of Clay.  While Nelson netted 10,000 more votes than Obama in Alachua, he also lost by 16,000 less votes in Clay.

The problem for Democrats, though, is that Clay is likely to become a larger share of the vote than Alachua with time.  Clay’s voter registration has been slowly increasing over the last eight years at a faster rate than Alachua.

Voter Reg Growth

Since 2006, Clay’s voter registration has increased from 113,000 to 136,000, an increase of 23,000 voters.  Meanwhile, Alachua has gone from 147,000 voters to 157,000, an increase of just 10,000.  Bradford, meanwhile, has hovered around 14,000 to 16,000 the entire time.

One of the problems for Alachua is that the student population of UF comes and goes, causing wild fluctuations in its registration.  In some years Alachua has actually fallen in registration as students graduate and move on while newer recruits are not registered to the county quick enough.  The bar graph below shows increases or decreases in voter registration over two year intervals since 2006.

Shifts in Voter Reg

Alachua saw increases in registration heading into Presidential years and then saw drops in the midterms.  No doubt part of this was OFA in 2008 and 2012 registering students and then midterms not seeing as much of a well-funded effort to update the registration of UF students who arrived in midterm years.  Clay saw its biggest registration increases in Presidential years.  However, before anyone thinks those gains were because of the Obama campaign, it should be noted almost all gains from 2010-2012 were from Repiblicans or NPAs and gains from 2006-2008 were more Republicans than Democrats.  Alachua relies on gains from students who eventually leave while Clay’s gains last; giving Clay a strong likelihood of eventually passing Alachua in registration.

In addition to registration gains, Clay has been slowly increasing its share of the vote in the district.


The bar graph above shows each counties share of the vote.  Alachua still has the largest share, staying above 50% in 2014 while Clay contributed 44%.  However, going back to 2006, Alachua had a higher share while Clay’s share was just under 40%.  The gap is slowing narrowing.  While Alachua increased in 2012, it fell back down in 2014, below 2010 numbers.  Bradford, meanwhile, always hovers around 5%.

If we were to subtract Clay’s vote share from Alachua’s to see the gap, we see that the gap has narrowed with recent elections.

Vote Share Gap

While Alachua managed to improve the gap in 2012 after a huge drop in 2010 (thanks to bad turnout in Alachua), the 2012 gap is still lower than 2008 Presidential and the 2006 midterm.  As Clay’s registration grows, strong Presidential turnout in Alachua will not be enough to keep the gap high.  I expect to see the gap continue to narrow with more elections.  The 2016 gap will probably be higher than 2014, but it won’t be as high as 2012.

With Clay slowly but surely creeping up on Alachua in registration and vote share, Democrats will only have a tougher time winning the district.

Possible Replacement Districts

The current boundaries of Senate District 7 are problematic in two ways. First, they are designed to ensure a Republican Senator.  The legislature knew of Clay’s steadfast Republicanism when they pared it with Alachua and knew the effect it would have on elections.  The other issue is that the district matches two communities very different from each other.  A suburban county like Clay fits in better with St. Johns or Duval.  Alachua fits in better with Marion County (home to Ocala) to the south.  In fact, it is easy to create a Senate district that links Gainesville to Ocala.

Gainesville Ocala District

Such a district shares much in common, linked by small communities and I-75.  Ocala and Gainesville would be the two main population centers of the district; with a Senator able to come from either area (most likely a Gainesville Democrat or Ocala Republican).  This would be a swing district that gave Obama around 50.7% of the vote and Nelson 56.5%.  The moderate suburbs of Ocala give moderate Democrats like Nelson the ability to run up a much stronger win that Obama does, giving Democrats plenty of opportunity to win the seat.

Another possible district continues the tradition of the old 14th, pairing Alachua with rural counties.  Alachua is fairly rural outside of Gainesville and its not crazy to pair it and other rural counties together.

Gainesville Rural District

In this case I try to split less counties, putting all of Putnam in and leaving Levy out.  Rural Marion gets added to even out the population and I leave sparsely populated Northern Columbia county out to be added to Baker or Hamilton counties.  This district would be influenced mostly by Alachua but give a Putnam politician a chance.  Like the old 14th district, it is a lean GOP swing seat that gives Obama 47% and Nelson 54%.  In this case, improving margins in the rural areas is key to victory for any moderate Democrat.


Regardless of what replacement district is offered, the fact is the current district screams of partisan intent.  The legislature, through its consulting firm, drew a district linking a moderately Democratic region to a hyper-Republican county to ensure no Democratic Senator came out of Gainesville.  The Jacksonville Suburbs of Clay County and the city of Gainesville have nothing in common are are not linked in any way except for the Senate District.  The goal is clearly to ensure a Republican seat; and that goal is achieved by putting a college town with one of the most Republican counties in the state.  The district serves as an important reminder that not all gerrymandering looks like it was drawn by a five year old.  Some of the cleanest looking districts can also have the most partisan of intentions.

How Florida Democrats fared in 2014′s Local Elections

In 2013, I wrote an article on Democratic Party Strength at the local level.  The article examined areas where Democrats were strong and weak on down-ballot races for county commission and constitutional officers.  It has long been my view that local elections are critical to the future of any political party.  Local elections allow parties to build benches for higher office.  In addition, local elections can be used to help measure party strength in different jurisdictions.  With the 2014 elections now over, I decided to take a look back at how Democrats fared at the local level.  While Florida Democrats did not take back the Governorship and six State House Democrats lost re-election; elections at the local level were not a Democratic defeat.  Despite losses further up the ballot, Democrats managed to take control of two county commissioner chambers and defend several targeted seats.   This article will look at some of those contest races and how they fared compared to the top of the ballot.

Local Control Heading into 2014

While there are many races for city government that took place in 2014; some of the biggest cities host their elections in high profile spring elections.  For 2014, I decided to focus on the county commission races, all of which, except in Duval, occur during state and federal elections.  Constitutional officers (Tax collector, Superintendent, Sheriff, Property Appraiser, Clerk of Court, Supervisor of Elections) are elected in Presidential years while county commission seats are rotated between midterm and presidential cycles (like Senate seats).  Therefore, this article largely focuses on the County Commission races and a few special elections for Constitutional Officers.

Before the 2014 elections, Democrats controlled 23 of the 67 county commissions.  This may seem low, but keep in mind Obama only won 13 counties in 2012 when he won with 50%.  Democrats controlled a large chunk of North Florida and rural counties that used to be reliably blue but have drifted away from state and federal Democrats.  Republicans, meanwhile, controlled several blue-leaning urban counties despite weakness at the top of the ballot.  The breakdown of county commission control as of December 2013 is below.

2013 County Commission Control2

Key Races in 2014

Heading into November 2014, there were several key races brewing for county commission seats and constitutional officers.  I profiled several of these races the day before the elections.  Please give it a read through for more background details on each race.  Some races that were not covered in that article will be covered now.  In some races, I went into much more detail about how local candidates performed compared to Democratic Gubernatorial Candidate Charlie Crist.  Any images for any race are uploaded in a minimized size.  However, if you click the image, it will load larger.

Alachua Commission and Tax Collector
Alachua had two major local elections.  Neither of these got coverage in my initial article due to an oversight.  Alachua Commissioners are elected at-large and Democrats held a 4-1 advantage on the commission. Crist won 56.1% of the county in the Governor’s election.  The lone Republican on the commission opted not to run again in 2014 and both parties fought for the seat.  In the end, Democrat Ken Cornell won with 56.7% of the vote, beating Republican John Martin.  In addition, the Tax Collector position was up in a special election for the last two years due to the death of the incumbent Democrat Von Fraser.  The race saw Democrat John Power and Republican Jon C. Costabile face off.  Both candidates worked in the Tax Collector office and pledged a positive campaign; claiming to be friends and colleagues.  Power won with 58.85% of the vote.  The votes for all three Democrats (Crist, Power, and Cornell) all correlate highly with each-other.  Both local Democrats performed slightly better than Crist, likely thanks to a lack of third party candidates siphoning off votes.

The scatter-plot below shows Crist % as they relate to Cornell and Power.  All three candidates had strong correlation.  The black line represents a perfect correlation with Crist’s percentages; which both local candidates come close to having.  As precincts became more Democratic, vote totals came closer together; while Cornell and Power outdid Crist (running above of the black line in the scatter-plot) in less Democratic areas.  The map below shows which precincts had the largest/smallest gaps between candidates.

Alachua Scatter    Alachua Difference

Bradford County Commission: Districts 2 and 4
Bradford had two competitive county commission elections.  Before the election, Democrats had 5-0 control of the commission.  Incumbent Danny Riddick was able to beat a Republican challenger who outspent him $3,000 to $13,000 with 53% of the vote.  Meanwhile, Incumbent Doyle Thomas lost to his Republican challenger with only 46.7% of the vote.  The elections took place in different districts of the county; both heavily backing Scott.  The result is Democrats holding 4-1 control of the commission.

Broward County Commission:  District 4
One big blow for local Democrats came out of Broward County.  Until 2010, Democrats controlled all 9 of Broward’s county commission districts.  Democratic Incumbent Ken Keechel lost to Republican Chip Lamarca in 2010.  However, the commission redrew the lines in 2012 and made the district more Democratic.  However, Keechel, seeking a rematch, fell short of taking back the seat 47.1% to 52.9%.  Keechle framed the races as Democrat vs Republican.  However, he was outspent by $300,000.  Keechel’s votes were heavily tied to Crist (as the scatterplot below shows).  However, he under-performed Crist in most precincts and especially in the heavily white, Republican coast.  The loss in county 4 is a blow for Broward Democrats. LaMarca is strongly positioned to run for congress or state legislature from his coastal district.  LaMarca benefited from Keechel having less funds and Keechel being seen as too much of a typical politician; he had run for a different county seat in 2012.

BrowardScatter2    BrowardMap

Calhoun County Commission:  Districts 2 and 4
In Calhoun, NPA commissioner Darrell McDougald was re-elected in a four-way fight with 34% of the vote against two other NPA candidates and a Democrat.  The Democrat came in last with only 17.5% of the vote.  In the other district, Democrat Dennis Jones won re-election of an NPA candidate with 63% of the vote.  The district was largely based out of the one heavily African-American precincts in Calhoun, the only one to vote for Charlie Crist.  Democrats shouldn’t be too excited about Jones’ hold since the seat was African-American.  They should be worried that a Democrat came in last in another district in county with overwhelming Democratic registration.

Columbia County Commission:  District 4
Democrats made another under-the-radar pickup, this time in Columbia County.  With Republicans previously controlled the commission 4-1 and only one seat being contested in November, Democrats could not gain control of the chamber. County Commissioners are elected on a non-partisan ballot, and party ID does not seem to play a major role in commission matters.  In an open seat left by a retiring Republican, Democrat Everett Phillips won the seat with 52.6%.  The district gave Rick Scott over 65% of the vote.  The chamber now stands at 3-2 Republican control.

Dixie County Commission:  District 4
Dixie County has made a stunning transformation in the last few election cycles.  The commission used to be heavily Democratic.  However, NPA candidates began winning the commission races and before long, the commission was 4 NPA candidates and 1 Democrat.  In 2014, that last Democrat ran for re-election as a NPA candidate.  Every candidate for the elections in district 2 and 4 were for NPAs despite candidates having the option to run with party affiliation.  Democratic Incumbent Jason Holifield won re-election as an NPA with 66.5% of the vote.

Escambia County Commission:  District 4
Local Democrats had hoped to make gains in the Escambia County Commission.  However, those efforts ran into the wall of Crist’s unpopularity in the panhandle of Florida.  Republican Grover Robinson managed to rack up 70% of the vote in his county commission district as the upper-income suburbs remained steadfastly Republican.  Crist did slightly better than Democrat Mike Lowery.  The two Democrat’s voters were fairly close together.  However, as precincts became less hard-line Republican, Crist did slightly better.


Franklin County Commission:  District 2
Democrat Cheryl Sanders was re-elected with 52.3% in a district that went strong for Rick Scott.  Sander’s win was in spite of Franklin shifting further right, rejecting Crist and Graham by more than 15 points despite its history of supporting Alex Sink for Governor and Bill Nelson for Senate.  This race serves as an example of a local Democrat surviving despite the top of the ballot shifting further red.

Gilchrist County Commission:  Districts 2 and 4 
Gilchrist’s commission became more Republican in 2014.  The commission was already 3-2 Republican, all elected at-large.  In November, Democratic Incumbent John Rance Thomas lost to Republican challenger Marion Poitevint 47.8% to 52.2%.  Crist, meanwhile, got just under 25% of the vote.  The Gilchrist tea party made it a point to oppose Thomas, and Thomas went out of his way to separate himself from national Democrats.   There were local controversies as well, involving business development and county administrative costs and salaries.  Both local and national politics likely contributed to Thomas’s defeat.  In another county commission race, Republican Incumbent Ray Harrison won re-election with 60.5% against a conservative Democrat, John Yencho.

As the scatterplot shows, Yencho and Thomas over-performed Crist.  Only one precinct gave all three candidates similar percentages.  That precinct has a large African-American population.  The remaining precincts gave the two democrats varying levels of support while never giving Crist more than 30%.  Both Democrats managed to keep their votes separate from Crist despite both losing.


Glades County Commission:  Districts 1 and 4
Democrats risked losing control of the Glades county commission in 2014.  However, Democrats managed to hold the District 4 open seat, and won back district 1; beating a Rick Scott appointee.  The result was Democrats returning to the 4-1 majority they held after the 2012 elections.  In district 4, Democrat John Ahern got 51%.  In district 1, Democrat Weston Pryor got 52%.  The raw vote difference between both Democrats was only 36 votes.  Despite having close vote totals, both candidates did better/worse in different precincts.  As the scatterplot shows, the two Democrats’ vote was far from a perfect correlation (which the red line represents).  The map below shows how Ahern and Pryor did compared to Crist.

Glades Scatter   Glades Compare

Gulf County Commission:  District 2
If I had to bet on which county commission would switch from Democratic to Republican control, I would have bet on Gulf.  The 3-2 Democratic commission seemed poised to go 2-3 as Democrat Ward McDaniel fought for re-election is a district that gave Rick Scott 61% of the vote.  McDaniel, however, managed to win with over 64% of the vote; even after being outspent 3-1.  The result is a testament to local politics trumping national.

Hillsborough County Commission:  At-Large
One of the heartbreaking local losses of 2014 was the Hillsborough County At-Large race.  Democrats were hoping to pick up this seat with Pat Kemp.  However, Republican Al Higginbotham narrowly won the seat by 2,000 votes.  Kemp managed to get 49.7% of the vote, but was likely weighed down by Crist getting a very weak 48.2% plurality win in the county.  15,000 less votes were cast for the county commission race than for Governor.  Kemp got 5,000 less votes than Crist.  However, Higginbotham got 7,000 more votes than Scott.  Higginbotham who served on the commission in a single-member seat, likely won some Crist voters and votes from 22,000 who votes for third party candidates for governor.

Crist and Kemp’s percentages were tied closely together.  Kemp outperformed Crist in most of the precincts.  Kemp’s over or under performing Crist had little to do with racial makeup of precincts.   In addition, precincts with a larger under-vote (people skipping the county race) were split among precincts that gave Kemp more support and those that gained less.  Kemp did under-perform Crist among the most heavily Democratic precincts (which were heavily African-American).  However, she beat out Crist in most of the heavily Democratic areas and other African-American precincts..

HillsKempCrist2    HillsboroughMap

Jackson County Commission:  Districts 2 and 4
In my view, one of the most consequential local elections of 2014 where in Jackson County, where a political earthquake resulted from the election of two Republicans to local office for the first time in Jackson County’s history.  District 2 Incumbent Democrat Edward Crutchfield lost re-election with 49.4% of the vote in a 2010 rematch against Republican Clint Pate.  In the seat left open by Jeremy Branch, Democrat Alvin Roberts only received 36.5% of the vote to Republican Eric Hill’s 50.9% (the rest going to NPA candidates).  Both districts were heavily Republican, with district 4 giving Crist around 28% and district 2 giving him 38%.  The county splits its precincts between county commission districts, making precinct-comparisons imperfect.  However, in the four precincts that were entirely in one specified district, the votes for Crist and the Democratic candidate for commission were not tied together.  Jackson County stood out as the last county in Florida to never elect a Republican to county office.  With 2014′s results, Republicans could begin to make a play for other commission districts and constitutional offices.  Jackson County should serve as a wake up call for North Florida democrats.  As Republican strength grows down ballot, upper ballot races in the region will become even tougher than they are right now and Congresswoman Gwen Graham cannot afford to see Jackson county trend further to the right.

Jefferson County Commission:  District 4
In an off-the-radar race, Democratic Incumbent Betsy Barfield won re-election against NPA Troy Avera 53.2% to 46.8%.  Jefferson leans Democratic thanks to the ancestral white Democrats, some state employees, and a large African-American population.  The county often gives competitive Democrats between 50% and 55% of the vote.  Local Republicans have been making inroads, taking Superintendent and nearly taking tax collector in 2012.  One NPA commissioner already serves with four democrats.  The district, comprising of only 3 precincts, narrowly voted for Scott.  Barfield’s vote was modestly tied with Crist, with Barfield over-performing him in all three precincts.


Lafayette County Commission:  Districts 2 and 4
Lafayette produced a mix bag for Democrats, who held control of the commission 5-0 before 2014.  In district 2, the Democrat won with 59% of the vote, while in district 4 the Democratic candidate only got 43% of the vote.  The winning Democrat won all 5 precincts, while the losing Democrat only got 1.  The result is Democrats retaining control of the commission 4-1 and both did much better than Crist.  However, the loss of one Democratic seat shows a continued Republican growth down-ballot.

Liberty County Commission and Superintendent
North Florida Democrats have been feeling the pressure from NPA candidates in recent elections. In addition, Liberty trended further right in 2014.  Both Gwen Graham and Charlie Crist got around 34% of the vote (Gwen getting 34.7% to Crist’s 33.7%).  The county had voted for Sink in 2010 and Nelson in 2012.  However, it remained stubbornly Republican in 2014, giving Graham the second weakest gain over Crist of any county in CD2.  Despite upper-ballot troubles, Democrats managed to hang on at the local level.  Both County Commission seats stayed Democratic against NPA opposition.  In district 2, Incumbent Dexter Barber won with 66.4% of the vote.  In district 4, James Sanders won the open seat with 56.7%.  In the Superintendent race, Democrat David Summers won with 35% of the vote against 3 NPA opponents.  Summers’ win is especially notable since he only spent $700 in the race.  For now, the Democratic near-monopoly at the local level holds with 5-0 commission control and 5 of the 6 constitutional officers.  However, top of ballot woes have gotten worse.

In the county commission races, the two Democratic candidate’s percentages were not closely tied to each other.  Barber, an incumbent ran stronger, but Sander’s did better in the two most African-American precincts (the top two in the northwest corner of the district).  In the superintendent race, Summer’s won despite an African-American NPA candidate winning the counties lone African-American majority precinct.  While local Democrats did very well in Liberty.  Their individual wins were not because voters cast ballots by party straight down the ballot, but rather individual dynamics in each race.  The scatter-plot shows that Barber and Sander’s percentages did not correlate, while the map shows how each Democrat won.

LibertyScatter  Libertymap

Manatee Commission At-Large
Manatee was a dark horse county that could have gone to Crist in the right environment.  However, as the Crist campaign faltered across the state, Manatee wound up giving Crist just 41.3%, worse than Obama had gotten in 2012.  As a result, Democratic commission candidate Terry Wonder got 40.3% after being outspent by a large margin.  In this suburban county, both candidates had their votes tied fairly close together.  However, in several suburban precincts, Crist or Wonder beat out the other.


Miami-Dade Commission:  District 8
Miami-Dade hosted its non-partisan elections for county commission in August.  Most races do not get excessively partisan in the county.  However, the August race for district 8 (located in the southern part of the county where Homestead lies), became extremely partisan.  Democrat Daniella Levine Cava defeated Republican Incumbent Lynda Bell 52% to 48%.  Unions and the local Democratic party heavily back Cava while Republicans pushed for Bell.  Her win resulted in a commission of 6 Democrats, 6 Republicans, and 1 No Party candidate.  The commission recently named Jean Monestime, a Democrat, its new chair.

Orange County Commission and Clerk of Courts, Amendment C
Orange county was bad for Democrats in 2014.  Three House Democrats in the county lost re-election and Crist got a weak 53%.  Democrats failed to pick up County Commission Seat 2, receiving only 41% despite Crist receiving 54%.  However, the race being non-partisan on the ballot likely did not help matters.   Democrats also failed to pass Amendment C, which would have made county commission races partisan.  In a cruel irony, another amendment that would make Constitutional Officers non-partisan did pass.  Democrats did manage to win the Clerk of Courts race.  Democratic candidate Tiffany Moore Russell got 52% to Republican Incumbent Eduardo “Eddie” Fernandez’s 48%.

Russell won the Clerk of Court race by sticking close to Crist in most precincts.  Both candidates vote’s strongly correlated with each other.  While Crist was white and Russell is black, Russell did not do especially better with African-Americans.  On the same notion, some areas where Russell did better than Crist were largely white.  Crist actually did 0.8% better than Russell in her county commission district.  Amendment C did correlate to smaller degree with Crist votes.  However, as the scatter-plot shows, it under-performed Crist and Russell in more Democratic areas.  That said, Amendment C had its strongest support in the African-American areas; with precincts voting Yes being almost exclusively African-American majority/plurality.  The second scatterplot shows that Amendment C support grew as African-American percentages increased.  However, Amendment C failed because it was weaker with white and Hispanic voters, and wasn’t strong enough with African Americans.  In the County commission race, Democrat Alvin Moore under-performed Crist across the precincts.  However, as the scatter-plot for the race shows, Moore’s percentages did increase as Crist’s did.  Despite not having a D next to his name, Moore and Crist still had a respectable vote correlation.  However, many voters clearly defected away for the county race.

OrangeScatter Orangemap OrangeBlackScatter  orangecountyscatter

Osceola County Commission:  Districts 2 and 4
Democrats succeeded in taking control of the Osceola county commission after picking up two seats in 2014.  Districts 2 and 4 were both held by Republican Incumbents facing strong Democratic challengers.  In district 2, Democrat Viviana Janer beat Republican John Quinones 51.2% to 48.8%. The district gave Crist 61% of the vote, which was too much for Quinones, a popular local Republican, to overcome.  Democrats will be glad to be rid of Quinones, who could have been a major threat for CD9 in future elections.  Meanwhile, in district 4, Democrat Cheryl Grieb beat Republican Incumbent Frank Attkisson 47.6% to 46.7% (with the rest going to a third candidate).  District 4 gave Crist 49.3% of the vote to Scott’s 44.5%.  Just last week, as the ban on same-sex marriage lifted, Grieb and her partner were the first same-sex couple to marry in Osceola County.

As the scatter-plot’s below show, Grieb and Crist had strong vote correlations in district 4.  However, in district 2, Janer and Crist’s vote totals were not correlated.  This could be largely thanks to Quinones’ strong popularity.

OceolaJaner   OsceolaGrieb


Palm Beach County Commission:  District 4
Republican Commissioner Steven Abrams defeated Democrat Andy O’Brien with 59.2% of the vote.  Charlie Crist won the district with 50.8%, however, O’Brien was unable to capitalize.  Abrams’ 2-1 money advantage no doubt helped.  O’Brien’s vote share had a modest correlation with Crist’s.  However, O’Brien under-performed Crist in nearly every precinct.


Pinellas County Commission:  At-Large and District 4
Democrats took control of the Pinellas commission by winning their third at-large congressional district in two years.  Before 2012, Democrats controlled 1 of the 7 commission districts (4 are in single-member districts, 3 are at-large).  Democrats initially held just the African-American majority district.  Then, in 2012, Democrats won the two at-large districts that were both up that year.  In 2014, Democrats were making a play for the last at-large seat.  Democrats nominated Largo Mayor Patricia Gerard, while Republicans nominated State Representative Ed Hooper.  Hooper actually beat the incumbent Republican commission in the August primary.  When the general election came, Gerard won 51% of the vote. The same day, Charlie Crist won 52%; actually a fairly mild showing considering Pinellas is his home county.  Gerard and Crist’s percentages were closely tied together.

Meanwhile, in district 4, Republican Dave Eggers won with 60.9% of the vote against Democrat Mark Weinkrantz’s 32.9% (the rest going to NPA candidates).  Crist won lost the district 45.8% to Scott’s 47%.  Weinkrantz percentages had some correlation with Crist’s, however, he under-performed Crist across the board.  The two scatterplots below shot how Gerard and Weinkrantz’s percentages related to Crist.  The map shows Crist’s win in Pinellas next to Gerard’s.

PinellasAtLarge   PinellasMapPinellas4Scatter

Suwannee County Commission:  District 4
Democrats picked up a seat for the Suwannee County Commission.  This race was off my radar, as Republicans maintained a 4-1 advantage and a pickup wouldn’t have shifted control.  Democrat Larry Carl Sessions won the district with 50.4% to Republican Harry K. Weaver’s 49.6%.  Every precinct in the district voted heavily for Rick Scott.  Sessions spent $9,000 compared to Weaver’s $3,000.  The result gives Republicans 3-2 control of the commission.

Taylor County Commission: District 4
Democrat Pam Feagle won re-election to the Taylor County Commission, winning 65.6% to Republican David Woods’ 34.4%.  Feagle sits in an overwhelmingly Republican district, but managed to win every precinct despite Crist loosing them all.

Final Result

As a result of the county commission elections, Democrats took control of two chambers; Pinellas and Osceola.  In addition, Republican-controlled Miami-Dade was put into a tie.

2014 County Commission Shift

The gains and losses are visualized in the map below.

2014 County Commission Gain

Democrats gained 8 seats while Republicans gained 5.  On paper this means Democrats made a net gain of 3 commissioners.  However, compared to where things stood after 2012, the math isn’t so simple.  First, Republicans made a gain in Putnam county when Rick Scott appointed a Republican to a Democrat-held seat and that appointee was not challenged by a Democrat for November.  This ended Putnam’s 5-0 Democratic commission.  In addition, the Democratic gain in Glades was to take back another Rick Scott appointee from a previously dem-held seat; so the gain just reverted things back to pre-2014.   This leaves Democrats with just one more seat than they held after 2012.  However, the independent gain in Dixie county was a Democrat retreating from his party to run as an NPA candidate.  That brings the increase in Democrats on the county commission to zero.  However, that seat goes to an increase in NPA held districts.  Unless I am missing any mid-cycle changes (possible considering there are alot of commissioners), the final party gains in terms of raw commissioners between 2012 and 2014 is

  • Democrats <Even>
  • Republicans -1
  • NPAs +1

Final control of county commissioners can be seen in the map below.

2014 County Commission Control2

Looking Ahead

Considering 2014 was such a bad year for Democrats across the board, the gains by Democrats in local office is a silver lining for the party.  Democrats took control of two major counties in the state; Pinellas and Osceola.  Democrats seizing control of all three at-large congressional seats in Pinellas over the last two years is especially notable and gives the party three prime candidates for David Jolly’s congressional seat.  Meanwhile, many Democratic-held seats in North Florida did not fall to the GOP.  However, Democrats suffering defeats in Jackson County may be a sign of future problems.  In addition, failures by Democrats to pick up seats in Broward and Hillsborough continue to give the GOP a foothold in counties they should not have.  That said,  Democrats were the clear winner in the local elections of 2014.  They gained control of commissioners and won critical and non-critical races.  In many instances, local Democrats were closely tied to races going on at the top of the ballot.  In other cases, Democrats outpaced or under-paced the top of the ballot.  Indeed the results were contradictory.  Some counties showed an amazing level of partisanship; yielding similar party results regardless of the race or circumstance.  In other instances, Democrats could win or lose regardless of party fortunes further up the ballot.   The urban elections such as Palm Beach 4 and Orange 2 showed a strong level of correlation between Democratic votes even if the Democrat lost. Meanwhile, the rural counties showed votes less tied together; with local dynamics mixing with national trends.

The notion that all politics is local is slowly giving way to increase nationalization of elections.  However, several counties in 2014 showed that local circumstance can still beat out national trends.  How long local Democrats and Republicans can hold off increased partisanship in the counties it has not already taken hold is something to watch in 2016 and beyond.

The North Florida Way: How Gwen Graham Won FL-02

November 4th, 2014 was a bad night for Democrats across the nation. Democrats fell short in several Gubernatorial elections, lost nine Senate Seats, and lost over ten House seats. Democratic fortunes in the South were abysmal. The Arkansas Governors Mansion went Republican, Senators Prior and Hagan lost their re-elections, and Senator Landrieu in Louisiana was forced into a runoff. Republicans also finally knocked off Congressman John Barrow and Democratic efforts to win seats Congressional seats in Arkansas, Virginia, and North Carolina came up short.

Democrats had bad news in Florida as well. Democrat Charlie Crist failed to win the Governors Mansion, Congressman Garcia lost re-election in Miami, and Democrats lost six seats in the state legislature. However, despite the bad results in Florida and in the South overall, there was one bright spot for Democrats; and it came from Florida’s 2nd Congressional District.

The District

The 2nd Congressional District is quite large, the largest geographical district in the state, and is made up of scattered rural voters and a few urban centers in North Florida. The district contains 14 counties, with the bulk of the population located in Tallahassee-based Leon County and Panama City-based Bay County.


The district is primarily Democratic, with only Bay County having more Republicans by registration. The district has a long Democratic heritage and loyalty. Most of the county commission and constitutional offices in the various counties are controlled by Democrats.  However, local party strength has not stopped Republicans from gaining the advantage further up the ballot.

2014 reg

The district is around 71% white, with the remaining voters largely African American and a small Hispanic and Asian population. For Democrats to win the district, they must overwhelmingly win the African-American vote and win between 35% and 40% of the white voters (depending on turnout). The African-American voters are heavily concentrated in the northern areas of the district.

2014 race


History of the District

Florida’s 2nd congressional district is your classic southern, ancestral Democratic district. It’s a district that is Democratic in registration and in control of local offices, but has become more Republican further up the ballot. Democratic strength is tied to heritage and loyalty. Rural residents feel that they are average working-class Democrats, and are moderate to conservative on most issues. They still identify with the Democrats of generations passed, but have grown more disillusioned with the national party. This disillusionment has not gotten any better with Barack Obama as President. While the numbers indicate race has played a small role in Obama’s weak showings in North Florida, the trend toward Republican support predates Obama’s election.  While Obama has not done well in the area, many white candidates have done poor as well.

The major liberal center for the district is Leon County, home to Tallahassee. Tallahassee features two universities, a community college, a large African-American population, and white suburbs filled with liberal white voters who are government workers or work with the University system. Gadsden County, the state’s only majority African-American county, also votes heavily Democratic in elections.

For decades, Democrats managed to hold onto the 2nd district. Democrat Allen Boyd held the seat from 1996 until 2010 when he lost re-election to Republican newcomer Steve Southerland. Southerland benefited from the red wave of 2010 as well as local fallout from Boyd’s tough votes in favor of the Affordable Care Act, Cap and Trade, and the Economic Stimulus Bill. With Boyd’s loss, the 2nd Congressional District was held by a Republican for the first time in over 100 years. That same night, Southern Democrats across several states lost their re-elections as well and several southern state legislatures became GOP for the first time since reconstruction. This Southern Democrat wipe-out was indeed an acceleration of the  southern realignment that has been happening in American politics for decades.

With the loss of Boyd, taking back the district proved over the next several cycles to be a monumental task. Former State Senator Al Lawson, who had almost beaten Boyd in a primary in 2010, challenged Southerland in 2012 but lost by 5%. Obama only got 46% in the district in 2012, winning just the urban Democratic Leon County, majority-African-American Gadsden County, and lean-Democratic Jefferson County.


Meanwhile, Democratic Senator Bill Nelson got 54% in the district in 2012, winning Liberty and Franklin Counties in addition to Obama’s wins, and getting better margins in the other counties.


As such, Nelson’s win serves as an important blueprint for Democrats in the 2nd Congressional District. While the district also voted for Democrat Alex Sink for Governor in 2010, the district typically has a federal bias when it comes to Democratic candidates. On average, Democrats running for state offices manage to get larger shares of the vote than those running for federal offices- the only exception being Nelson.


The main reason for this federal bias seems to be rural voters in the district are hesitant to send a Democrat to Washington. This hesitation has grown as federal elections become less about local issues and instead become proxy wars for the national parties and their leaders.  The counties listed below are the ones which have demonstrated a bias against federal candidates on average.  The chart also shows the average percent federal Democrats under-perform their state counterparts.  More analysis on the federal versus state issue in North Florida can be found here

Calhoun -2.77%
Franklin -3.84%
Gulf -0.34%
Holmes -4.46%
Jackson -2.37%
Jefferson -2.15%
Liberty -5.60%
Madison -0.07%
Taylor -1.81%
Wakulla -4.51%

Nelson’s 54% in the district certainly indicates that Democrats still have a chance in the area, but the path to reclaiming the seat was still considered a daunting task leading into this cycle.

2014:  The Emergence of Gwen Graham

Heading into 2013, there was a great deal of speculation on who would run, or should run, against Southerland in 2014. One daunting issue was that 2014 would be a midterm during Obama’s second term, which historically favors the party out of power. Speculation centered around several local officials in the Tallahassee area. However, one name began to emerge in early 2013: Gwen Graham.

Graham, an attorney who worked for the Leon County School District, began to be floated as a possibility early on. As the daughter of former Governor and Senator Bob Graham, a legend in Florida politics, she seemed like a strong choice for office. However, Graham was fairly unknown in the beginning. There was also a very real threat of a primary challenge, particularly from Al Lawson, that would have likely stopped the candidacy in its tracks.

However, Graham started the campaign at a fever pace, going to Democratic events, meeting voters, and winning over activists and average voters early on. The campaign brought on a talented campaign manager with a strong fundraising background, Julia Gill Woodward, 18 months out from Election Day. The campaign raised over $300,000 in its first fundraising quarter, immediately garnering major attention in the national press and establishing Graham as a formidable candidate for the district. While critics tried to claim Graham was trying to ride on her fathers name, those critics were silenced by her strong
work ethic, impressive fundraising power, and personal likability. Graham, committed to running a true 14-county campaign, wooed crowds in the summer of 2013 across the entire district.

Meanwhile her campaign continued to build a powerful finance team to raise money through the fall of 2013. The campaign hired Elizabeth Davis as its Finance Director and Brendan Olsen as its Deputy Finance Director. With this team in place, the campaign continued to raise hundreds of thousands of dollars at an impressive rate.

In early 2014, the campaign began to build a field team as well, bringing on Jenn Whitcomb, the Obama Regional Director for the region in 2012, as the Field Director. With this unusually early investment in Field, the campaign began registering voters, collecting petitions for ballot qualification, and grassroots communication with voters. The field program quickly established itself as one of the best operations in the nation.

The campaign continued to grow into a juggernaut in North Florida throughout the year. The field operations were poised to surpass that of the Obama campaign of 2012 within the district. Graham’s personal commitment and discipline as a candidate allowed the campaign to continue raising money at an astonishing rate, out-raising Southerland nearly every single quarter. The campaign also brought on Eric Conrad, a seasoned communications guru in Florida, as Communications Director. Graham accomplished early on perhaps one of the hardest tasks for first-time candidates: bringing on smart strong hires and then trusting them to execute.

The campaign and its communications staff came up with what may be the best campaign slogan of the cycle: The North Florida Way. By summer of 2014, when Gwen Graham decided to go on air (and had the airwaves to herself for several months), the slogan was used in every spot. Graham promised to work across the aisle and derided the partisan nature of both Congress and Congressman Southerland. She promised to work for the betterment of the community and to always represent North Florida’s interests in Congress.

This tactic was brilliant. By going on television early and ensuring the voters she was “one of them,” Graham had helped inoculate herself to the eventual attacks that would come from Southerland and his allies, as well as subdue the district’s bias against electing federal Democrats. The eventual attacks that came were predictable– that Graham was just a Tallahassee liberal who would support Barack Obama and Nancy Pelosi. Graham pushed back, saying she would not back Pelosi for speaker and argued that new leadership was needed in both parties. Graham also hit Southerland for his partisan voting record, for not voting to end of the Government shutdown, and his vote against the Violence Against Women Act, after he claimed to have supported it.

Southerland did himself no favors when it was revealed he had organized a ‘mens only’ fundraiser for his campaign. When confronted on the issue, Southerland laughed off the concern and asked the reporters if Gwen had, and I am quoting here, had ever gone to a “lingerie party” before. The national press seized the story and Southerland was painted as a sexist. As election day got closer, the national press declared Southerland one of the most endangered GOP incumbents in the nation. Republicans quoted off the record were furious that the seat was in play at all, much less the campaign’s responses to Graham’s attacks. By the time election day came around, everyone knew it would be close.

The Results

Election night was a roller-coaster when it came to Florida’s 2nd District.  Gwen racked up a strong margin as the first returns came in from the eastern region of the district. Leon County counted quickly and at one point Graham had 55% of the total counted vote in the district.

The concern, of course, was the western region and the rural counties. The Graham campaign didn’t need to win the rural counties, they just needed to not get killed in them and mitigate Southerland’s wins. The biggest concern was Bay County, stationed in the west. Bay was Southerland’s home base and it only gave Lawson 27% when he faced off against Southerland in 2012. In fact, in 2012 Bay’s raw vote margin for Southerland had cancelled out the margin Lawson got from Leon (despite Leon’s population being higher). The Graham campaign had hoped to both increase its Leon County margin and narrow Southerland’s Bay County margin. While there were many different paths for the Graham campaign to achieve victory, getting 30% in Bay, 65% in Leon, and slightly better percentages than Lawson and Obama in the rural counties was the most likely strategy for winning.

As the western counties began to come in, it was clear Graham had a real chance of doing just that. Bay’s turnout was only 51%, while Leon’s was 57%. By the time Bay really started reporting, Leon was largely in and was giving Graham 65% of the vote, just where she needed to be. Bay was hovering around 30% for Graham. With the higher support and lower turnout, the Bay margin failed to erase the Leon margin. Graham netted 32,000 votes from Leon, while Southerland netted 22,000 from Bay. This left Graham with a net 10,000 votes. Graham netted an additional 8,000 votes from Gadsden and 800 more from Jefferson.

As the rural counties trickled in, voting for Southerland but not by the margins they had in 2012, the lead began to slowly shrink.  Graham’s improvements in the rural counties made it so that they could not completely erase her lead, and with only a handful of precincts left in rural counties and Graham still leading by 3,000 votes, it became clear Graham would win.  When the final numbers were in, Graham had won by just over 2,200 votes.

Florida Congressional District 2 Results

Map Note:  The margin factors in a write-in candidate, and Jefferson precinct results are election day only.

Graham’s win served as one of the few bright spots for Democrats on election night. In Florida, Democrat Charlie Crist failed to beat Rick Scott in the Governors Race, and lost the 2nd district with only 46% of the vote.  Both candidates won the three Democratic Counties, but Graham had better percents than Crist in every county.

Cd2 Compare

Graham over-performed by 7 points in Bay County.  This is especially notable considering that Graham’s opponent hailed from Bay County.  Crist under-performed Graham across the board, only doing better in select precincts.


Graham’s near-universal improvement over Charlie Crist is also notable for the fact that as a federal Democrat, she had the disadvantage.  The gap shows the strength of Graham and the weakness of Crist.

Crist’s under-performance of Graham was not due to people skipping the Gubernatorial race either.  In fact, more voters were cast in the 2nd district for Governor than for Congress.  Around 4,000 more votes were cast in the Gubernatorial Election in the district.  This number is slightly inflated by the fact that Holmes and Madison split precincts (in districts 2 and 3), so in those precincts, the numbers for Governor are for the whole precinct, not just within the 2nd District.  However, even taking those precincts out, slightly more voters were cast for Governor.  The map below shows the percent of the voters who cast votes for congress as well as for Governor by precinct.

2014 Undervote

As the map shows, a vast majority of precincts had vote totals for congress than were at least 98% of those cast for Governor, indicating few left either race blank.  The green precincts show where more voters were cast for Congress than for Governor. There was no clear partisan pattern in whether more votes were cast for Congress or for Governor.

Despite Graham’s improvement in the margins, there were still many Democratic areas that rejected both Graham and Crist. The map below shows precincts over 50% in Democratic registration, but voted largely Republican.

2014 Southern Democrats

Graham’s win is amazing, but it hasn’t stopped the red slide of Southern Democrats in North Florida just yet.

Looking at a precinct map of Graham’s win and Obama’s loss from 2012, the precinct map is largely the same.  Graham only won a few precincts that Obama lost.  Meanwhile Obama won two precincts Graham lost.

2014 Loyalty

Map Note:  It should be noted that the Madison precinct that voted for Obama but not Gwen is a split precinct, and its Presidential (and any Gubernatorial results) represent the whole precinct, not the section that was in Congressional District 2.  Any comparisons in the 5 or so split precincts in Madison or Holmes are imperfect for this reason.  

The key for Graham wasn’t winning areas the President had lost.  It was simply doing better than him across the board, regardless of the precinct being red or blue.

2014 Gwen Obama Compare

Graham over-performed President Obama across the board in the district. She only under-performed him in scattered precincts.  The precincts where Obama did better were 37% African-American. The loss of support can be attributed to the white voters moving more Republican or higher Republican turnout in those areas.   Graham actually over-performed the President in many African-American areas, particularly in Leon County.  Graham did better than Obama in Tallahassee’s southside, beating the President in precincts that were over 90% African-American.  One particular note, Graham got a higher share of the vote than Obama had in Florida A&M University, a largely African-American college.

Graham’s biggest bases of improvement from Obama were with white voters.  As precincts became whiter, her increase from the President’s percent grew.

White Gwen Gain

Still, Graham’s over-performance with African-Americans was something even Bill Nelson could not do.  Nelson won the district by limiting losses with white voters.  However, he did not get as high of percentages as Obama in the African-American areas.  While Graham did not do as well with white voters as Nelson had, she did better with African-Americans than both Nelson and President Obama.

The scatter-plot below shows Graham (blue), Nelson (red), and Obama (green) percentages by precinct and how they fluctuated as precincts became whiter.  The left end are the least white precincts (essentially African-American) while the right side are the whitest.  As the plot shows, Graham and Obama started off higher than Nelson in the African-American precincts.  As precincts grew whiter, support for all candidates fell.  Graham and Obama fell at a quicker rate than Nelson.  However, Graham did not fall as far as the President.

graham nelson obama white

Graham starting off high with African-Americans and not falling to far with whites were instrumental to her win.

Graham’s improvement of Nelson with African-Americans can be seen in the map below showing where Nelson or Graham performed best.  Nelson, who has long been popular in rural Florida, did better than Graham in most of the rural parts of the district.  However, Graham did better than Nelson in African-American precincts and in several white suburbs in Leon County.  Graham improved over Southerland in swing Leon suburbs like Betton Woods and Waverly Hills, in addition to Republican suburbs like Killearn Estates and Southwood.

Graham and Nelson

Nelson’s strength in the rural regions of Florida date back over a decade, so his over-performance of Graham shouldn’t be surprising.  Nelson’s stronger over-performing areas were in Taylor, Franklin and Liberty; in very low populated precincts.  Meanwhile, Graham’s improvement over Nelson in the well populated suburbs were instrumental in increasing her needed margins in Leon County.

Where Graham succeeded, Crist failed.  Despite being a former Republican, Crist ended up as “just another Democrat” in North Florida.  Crist over-performed Obama in the eastern end of the district, but lost ground further west.  This was especially notable in Bay County.

2014 Crist Obama Compare

Crist lost ground to Obama in all of Bay and in scattered rural precincts.  In addition, he lost ground to Obama in much of the African-American community.  Crist’s losses with whites in Bay and African-Americans throughout resulted in him getting a lower percentages than Obama had in 2012.

Crist’s percentages-by-precinct closely correlates with Obama’s. The red line in the scatterplot below would represent a perfect correlation (Crist gets a certain percentage, Obama has exact same). Nearly all the precincts, and the yellow line representing Crist’s trend, falls close to that perfect correlation line. Overall, whether it was under or over-performing the President, the differences in percentages between the two were minimal.

Crist Obama Scatter

To show again how Graham did better at separating from Obama’s weaknesses in the 2nd district, lets look at a scatter-plot of Graham and Crist percentages compared to Obama’s.

The plot below shows Obama’s percentages by precinct from right to left, and Crist’s (red) and Graham’s (blue) up and down.

Graham Crist Obama

The plot shows a great deal of correlation for both candidates.  Overall their support was closely tied to the President’s support.  However, the big difference is Gwen got a higher percentages in the weak Obama precincts (the bottom left).  Precincts where Crist and Obama might only get 20%, Graham would manage 30%.  Graham over-performed Crist and Obama in the redder precincts.  The gap narrowed as precincts became more Democratic, but Graham managed a narrow over-performance until the they reached the most Democratic precincts.  Those gaps in the red areas were very important for Gwen’s win, as the race came down to margins.

One final piece of analysis in the second district revolves around Nan Rich.  Rich, a liberal, Jewish, former State Senator, did best in the conservative panhandle counties thanks to opposition to Crist.  A vast majority of voters didn’t know who Rich was and their vote for her was a reflection of the conservative Democrats’ opposition to Crist.  Rich got 31% in the second district, higher than her statewide total, winning several precincts.

2014 Rich Crist Precincts 2nd

Rich’s precincts where 42% Democratic and over 85% white.  These were not liberal enclaves, but rather areas filled with very conservative Democrats who had grown disillusioned with the party and voted for the unknown Rich as a protest.  Crist got 20% in those precincts compared to the 26% Graham got.  In fact, as precincts supported Rich by a higher margin, their support for Graham or Crist went down.

2014 Crist Graham Rich

Graham managed to bleed support less than Crist in the heavy Nan Rich precincts.  This stands out as another example of Graham outperforming Crist in the heavily conservative areas.

Field Wins Elections

Field doesn’t get nearly as much attention or money as TV and direct mail do, but it is a critical component of any well-run campaign. While field got more attention than usual in the course of the Obama campaigns, it is still not the fixture of others. Most campaigns use field operations, but often with limited financial resources, or outsource entirely. The former was seen in the Crist Campaign this cycle, where field programs outside of southeast Florida were lacked resources and goals. In fact, the Crist campaign’s lack of presence in North Florida resulted in them losing the district despite Scott’s immense unpopularity with state employees, a major voting block in Leon County.

The Graham campaign, meanwhile, had one of, if not the, largest congressional field programs in the country.  The program included efforts focused on volunteer recruitment, voter registration, persuasion and get out the vote drives during Early Vote and on Election Day. In addition to that, Graham’s campaign focused Vote By Mail, signing up a record number of supports to use that method of voting in CD2.  The campaign was successful in registering over 11,000 new voters in the district.   The campaign’s hiring practices were also commendable.  The field program hired staffers based on skill, knowledge, and how they reflected the communities they would be responsible for working in, with a great deal of diversity throughout the field hierarchy.  The campaign maintained a presence in every county of the district and left no area neglected.  The campaign’s dedication to reaching all voters and community involvement can be seen in Graham doing better in the African-American community than even President Obama.

When looking at the final margins for Graham, just over 2,000 votes, it is clear that if any one piece of this massive operation had gone wrong, Southerland would have prevailed. But the field campaign was a well oiled machine that worked to get out Democratic voters as well as persuade rural whites. Without the field program, Leon’s turnout would have fallen lower. The campaign’s early investment in Bay also ensured Graham got 7 points higher than Crist.

The Graham and Crist numbers truly highlight the importance of field.  Both candidates had good name recognition by election day, ad both has large sums of money on TV and mail.  Crist was known in the area, had won the district in 2006 as a Republican, and had won Leon as an Independent in 2010.  Both Graham and Crist were running against flawed GOP incumbents. Graham’s margins over Crist, and path to victory, was thanks to the well-oiled machine that was its field program.  That cannot be stated enough.


Gwen Graham’s win is important for two reasons.  First, it is a testament to how good campaigns can matter.  Anyone who has worked in campaigns knows the sting of running a good campaign but falling short due to outside circumstance.  The Graham campaign, however, managed to defy the Republican year and the Republican lean of the district to be one of the few Democratic pickups in 2014.  Second, the Graham win has slowed the tide that has been growing stronger in the 2nd congressional district.  The region is still trending Republican.  Crist’s loss in the 2nd district is a reflection of both its lack of investment and a Republican trend.  Republican gains in local offices in North Florida on election night are a reflection of the continuing realignment in the area.  Graham’s win stands completely counter the the growing Republican sentiment in North Florida.  It stands as a testament to what can be accomplished with a strong campaign.  She will need an equally strong operation in 2016.  Demographics and electoral trends can be hard to overcome.  However, as Gwen Graham shows, good campaigns matter.