I had an email question about the Gainesville Mayor’s race from 2013.
“How did a tea-party labeled Republican win in dark blue Democratic Gainesville?”
First some background and context on the Gainesville Mayor Race.
Context for 2013’s Mayor Campaign
Gainesville’s elections are officially non-partisan; however, that does not prevent both parties from playing prominent roles, essentially striping away the illusion of nonpartisan elections.
Gainesville elects its Mayor by popular vote. Gainesville has a weak mayor system, with an election held every three years, in or around the March/April time frame. Held concurrently is a Gainesville district race in one of the most liberal (containing the infamous Duckpond) districts in Gainesville – District 4 .
Base Rate for Success for Republican Candidates
Gainesville has popularly elected a Mayor since 1998, since that time there have been 6 Mayor elections. Of those elected, 2 have been Republicans. A base rate of success for Republicans in Gainesville Mayor’s elections is 33%.
Some argue the first elected Gainesville Mayor’s race was essentially uncontested and the modern injection of the political parties didn’t happen until the following election where the incumbent Republican lost; therefore the base rate should be 20%.
In 2013, a Republican Ed Braddy ran against an incumbent Democratic Mayor, Craig Lowe. (My man has is own Wikipedia page). According to his Wikipedia page Craig Lowe became Mayor after he won “a run-off election on April 13, 2010, by a margin of 42 votes (which held through an automatic recount) Lowe became Mayor-elect of Gainesville. He was sworn in on May 20, 2010, becoming the first openly gay Mayor of the city.”
Craig Lowe’s first term was highlighted by his successful handling of the infamous “Pastor” Terry Jones and the Dove Outreach’s plans to burn a Quran on the 9/11 anniversary. Mayor Lowe was widely praised for his part handling a volatile situation.
Criag Lowe’s first term was also highlighted by charges of favoritism in hiring Lowe’s campaign manager as a city employee, Lowe’s defense of the GREC biomass energy plant and electric rate increases, and Lowe’s heavy handed moves to limit citizen input into government proceedings.
Heading into the election, Craig Lowe was facing headwinds for re-election; however, in Gainesville it takes more than a headwind to stop the Democratic tidal-wave.
Ed Braddy was a previously two-time elected Gainesville commissioner from a district (district 2) in the North West part of town. The district has a history of performing well for Republicans; however, Braddy was popular and well liked in this district.
Braddy was well known for being a Republican/conservative voice on a commission dominated by Democrats.
During his second term in 2006, Ed Braddy was arrested for DUI. Braddy spent time in treatment, plead no contest, was sentenced to one year supervised probation and a $250 fine. Braddy’s drivers license was suspended for six months, and Braddy was ordered to complete 50 hours of community service including five hours with a victim impact panel. A civil traffic infraction against Braddy was dismissed. (This becomes important shortly.)
Once Braddy reached term limits in his district seat, Braddy was a on-air host for a local radio talk show that was popular with conservatives the two years leading up to his election.
General Election for Mayor
The campaign for Mayor was initially a 6 way race between the incumbent Mayor, 2 former city commissioners, 1 business leader and two others. The race was held 3/21/2013 with a 15% turnout.
The candidates receiving the two highest vote totals moved on to a run-off to be held approximately a month later on 4/16/2013.
The results of the General election are as follows:
These results would indicate trouble for any incumbent. Even in a 6 way race any incumbent should do better than 37% – especially when the District 4 election was held at the same time.
Other things of note, The District 4 election did not goto a run-off and Commissioner Henry is an African American and we will explore the significance of this later.
Run-off Election for Mayor
The run-off election was set between Ed Braddy (Rep) and Craig Lowe (Dem).
The run-off began with a bang when Mayor Lowe was immediately busted for DUI on March 20 after crashing his car. Mayor Lowe issued a widely criticized statement after the arrest. On April 2, Mayor Lowe and the State Attorney agreed to “a deferred prosecution agreement that resolves his DUI case. According to news reports, Lowe was required to complete all requirements for those convicted of DUI, monitored for 18 months and gave up his driver’s license for 14 days. The program also included 50 hours of community service and an alcohol evaluation.
The candidates’ DUIs became an issue in the campaign as well as biomass, transportation, and Ed Braddy’s conservative talk radio. The campaign also contained a last minute mail attack from the Democratic Party affiliating Ed Braddy with the Tea Party. The campaign also had a highly controversial open letter from 19 formerly election Democratic officials published by several news outlets that endorsed Craig Lowe while highly critical of Ed Braddy calling him “a relentless, inflammatory, and often dishonest critic.”
In the end, on April 16, 2013, Ed Braddy won a decisive victory with 55% of the vote, becoming the second elected Republican Mayor in Gainesville’s history.
Turn out increased from the general election of 15% to to 16.2%
Analysis of the 2013 Mayor’s Race
We begin our analysis at looking at the differences between the electorate’s composition from the General election and the Run-off Election. Data Notation
GnR – voted in general but NOT run-off
G&R – vote in BOTH general and run-off
RnG – voted in run-off but NOT general.
We then turn our attention to turnout. The following three maps show the location and intensity of voters by precinct who 1) Only voted in General Election, 2) Voted in Both General and Run-off Election, and 3)Voted in Run-off election only. (click map for larger view)
Hypothesis #1: Ed Braddy was elected Mayor on the strength of Republican/TeaParty/Conservative votes alone.
It is safe to say, 50% or more of Ed Braddy’s support came from non-Republican voters.
In the Run-off Election, Ed Braddy received 7,267 votes. Even if we give the impossible but theoretical 100% of the Rep and Other vote totals to Braddy, 5,139 votes, Braddy must have received a minimum of 2,128 votes from Democrats.
If we give the impossible but theoretical every single Republican vote to Ed Braddy, 3693 votes, 3574 votes came from other than Republican voters.
Hypothesis #2: Due to the Lowe’s DUI, Lowe’s voters didn’t show up in the run-off to vote for him, instead choosing to stay home.
There are some marginal findings to support this.
When we explore a correlation of Lowe’s run-off percentage and the % difference in turnout between the general and run-off elections by precinct, we observe a slight negative correlation.
The real, statistical significant finding is the opposite correlation. When we explore the correlation between Ed Braddy’s run-off percentage and the difference in turnout from the General to the Run-off election, we see a statistically significant positive correlation.
Meaning, there is evidence for a slight stay-at-home effect for Craig Lowe, but there is more evidence for a surge in turnout for Ed Braddy in the Run-off.
Maps showing candidates % of vote by intensity (click map for larger view)
We observe a decline in the % of Democrats as a % of the Voters from the General to the Run-off election.
We observe a notable increase in % of Republicans as a % of Voters from the General to the Run-off election.
Hypothesis #3: African Americans didn’t vote for Craig Lowe.
There is some evidence to support this hypothesis. Democrats traditionally do well with African-American voters.
With Commissioner Henry on the ballot in the General Election, Craig Lowe did not perform well in heavy populated African American precincts.
With Commissioner Henry’s defeat in the general election, there was no African-American candidate on the ballot during the run-off election.
In looking closely at the race break down from the General Election to Run-Off election, we observe 291 fewer African American’s voting in an election that increased in over-all turnout. African American’s decreased as a percentage of turn out from 15% to 11.4% (-3.6%) of the electorate.
You can observe in the maps above between the General and Run-off elections.
Hypothesis #4: Students didn’t vote.
Let’s rephrase this to young voters under 24 didn’t show up for the run-off election.
We can see the largest number of under 24 voters showed up for the first, general election, but didn’t return to vote in the run-off.
We observe 698 twenty-four and under voters voting in the general election, but only 418 voting in the Run-off election. Young Voters decreased as a percentage of turn out from 6% to 3.3% (-2.7%) of the electorate.
The student precincts are precincts where Craig Lowe did well in the General election and past Mayoral elections.
Hypothesis #5: Many new voters who traditional don’t vote in city elections surged to the polls.
|City Score||0||1||2||3||4||Grand Total|
Scoring all voters by two scores: City Score (number of times voting in last 4 Mayors campaigns) and County Score (number of times voting by last 4 county wide races), we see that 2049 or 16% of the run-off voters were casting their first vote in a city election.
So, let’s return to our original question: “How did a tea-party labeled Republican win in dark blue Democratic Gainesville?”
Inquissima haec bellorum condicio est: prospera omnes sibi indicant, aduersa uni imputantur
Rough Translation: This is an unfair thing about war: victory is claimed by all, failure to one alone
Tacitus, Agricola 27:1 (written ~ 98AD)
As with any campaign, there is no single answer for victory or defeat.
- Did Ed Braddy win solely because Craig Lowe drove his car through a stop sign while drunk three weeks before the election?
- Was Craig Lowe’s DUI more damaging than Ed Braddy’s DUI due to recency?
- Were people fed up with electric rate increases due to biomass?
- Did people take great offense at an open letter from 19 officials telling them who to vote for?
- Were voters fed up with the paving plans for streets (16th & 8th Ave) and / or reduction of traffic lanes?
- Did Craig Lowe’s sexual orientation conflict with Christian African American Voters?
- Did Craig Lowe not invest enough time with African American voters?
- Did the students show up in force for the Mayor’s first election in the 2010 because of the remnants of President Barack Obama’s student organization? Then fade away for the Mayor’s second election due to attrition, but not yet being revved up for the President’s reelection?
- Is there a growing two Gainesvilles with the split growing between the urban core and suburbs?
It is fair to say that in order for a candidate to overcome a base rate of success of 33%, everything must work in concert with another in a perfect manner.
Some of my Democrat friends say plainly, “Ed Braddy got lucky, being at the right place at the right time.” Hogwash. Mayor Braddy is a qualified, skilled candidate who ran a good campaign that assembled a rough coalition of people who wanted change in the Mayor’s office that was able to seize when fortune presented itself.
In summation, I start with the belief that elections are referendums on the incumbent foremost. I think this election can be boiled down to the desire to get rid of the incumbent Mayor clearly outweighed the desire to save him even among those in his own party.
However, the very item Braddy owes his success to is also his challenge – especially for Braddy’s re-election.
The question is: Can Ed Braddy hold together dissimilar parties? Mayor Braddy is a highly skilled and practiced politician, but holding the coalition together without the foil of an unpopular incumbent and outside the context of the last election will be difficult at best.
Does Gainesville regress to the mean? This is to be determined.
As astute reader will notice that the totals for turnout in the data analysis do not match the actual turn out numbers on election day. This is due to only active voters being provided by the local Supervisor of Elections.
Active voters “exclude those who have moved out of county, been dropped during file maintenance, deceased, etc. Public lists also exclude voters with records exemptions on file to avoid disclosing their information directly or indirectly. “
General = 12094-11555 = 539 (.045)
Run-off = 13278 – 12644 = 634 (.048)
4.5% of the voters who cast ballots in these elections is missing from this analysis.