International Case Study


Clarity Helps Guide Colombian President to Reelection Victory

SCROLL DOWN

International Case Study


Clarity Helps Guide Colombian President to Reelection Victory

Model dashboard created for the Santos campaign

For the first time in Colombian history, Clarity Campaign Labs applied modeling and analytics to the nation’s presidential election and is proud to have helped guide candidate Juan Manuel Santos to victory. In the June 2014 runoff election, President Santos won with 50.9% of the vote against opponent Oscar Ivan Zuluaga, who garnered 45% of the electorate. This victory will allow Santos to continue critical peace talks with Fuerzas Armadas Revolucionarias de Colombia (FARC) rebels to end a civil war that has been raging for half a century.

The initial round of voting, in which President Santos, with only 25.7% of the vote, took second place behind Zuluago, had a 39% turnout rate, the lowest Colombia has seen in 20 years. With the help of Clarity’s analytics, the Santos campaign was able to increase both support and turnout for the runoff election. Using previous election results, census data, geographic economic statistics, and contemporary polling the Clarity team was able to create comprehensive support models for President Santos at both the municipal and polling location level of geography.

The models on the polling location scale essentially predicted strong support in the areas surrounding each polling location for the five largest cities in Colombia, Bogotá, Medellín, Baranquilla, Santiago de Cali, and Cartagena, proving very important for the campaign’s get out the vote (GOTV) efforts. After the first round of voting, Clarity updated the support models to enhance the accuracy for GOTV targeting for the runoff election. In the runoff, 47.89% of the electorate voted, an increase of almost 9% over the first round, demonstrating how data and analytics can help shape the outcome of a race.

President Santos won by less than a million votes with 50.95% of the runoff election vote. The model proved so accurate that it missed the true results by less than 114,000 votes in a runoff election that more than 14.7 million people cast a ballot.

Equality Case Studies


Since we were founded in 2012, Clarity has been proud to work with organizations around the country fighting for LGBT people and our team has been working on modeling and analytics for the freedom to marry since 2009.

Equality Case Studies


Since we were founded in 2012, Clarity has been proud to work with organizations around the country fighting for LGBT people and our team has been working on modeling and analytics for the freedom to marry since 2009.

Fair Pocatello

In May of 2014, we partnered with Fair Pocatello to provide data and analytics at an affordable scale for the effort to defend Pocatello, Idaho’s anti-discrimination statute from being repealed at the ballot box. The opponents of the statute placed the repeal effort on Primary Election day in this heavily Republican town, and supporters had an uphill climb in the conservative, heavily Mormon city.

Fair Pocatello had neither the budget nor a large enough city to make custom modeling possible, but we were able to leverage our existing data to identify universes of likely supporters to turn out, likely undecided voters to persuade, and likely opponents to avoid. In the end, the statute was saved by 80 votes out of almost 10,000 cast – less than one percent of the vote.

New Jersey United for Marriage

In 2013, in anticipation of a statewide ballot effort for marriage equality or a major legislative effort to override Gov. Christie’s veto, we built support scores to predict the likelihood of Garden State voters to support marriage equality. The courts made those efforts less relevant, happily, but the scores are available to NJUM and their partners for future efforts.

Basic Rights Oregon

In 2013, Clarity worked with Basic Rights Oregon, who were fighting for marriage equality in their state, to build a series of models predicting support for equal marriage. While BRO was interested, of course, in finding supporters of full marriage equality, we discovered through our work with them that we could better identify persuasion targets by asking the questions differently. Instead of simply asking a yes-or-no question on marriage equality, we gave respondents three choice – full equality, civil unions, or no recognition – and built models to predict how likely Oregonians were to support each of those three positions. Voters who scored the highest on the civil unions score were the most persuadable on full marriage equality. Of all voters who started as unsupportive, they were the ones who could be brought around, and therefore formed the core of BRO’s persuasion efforts. 

Other experience

Our team has also created a variety of LGBT voter contact models and advised several other marriage equality campaigns going back to 2009 through our experience with ISSI, America Votes, and State Voices:

  • Maine 2012: Models for second campaign for marriage equality
  • Minnesota 2012: Models to defend against anti-marriage amendment
  • Montana 2011: Model among multiple issue scores for the c3 table
  • Oregon 2010: Models to prepare for full marriage campaign (updated project described above)
  • Washington State 2011: Models to prepare for full marriage campaign
  • Washington State 2009: Vote Yes on Ref 71 domestic partnerships