Clarity's own Dan Castleman, Co-founder and Director of Analytics, has been featured in the recently published eBook, Data and Democracy, available for free download (via O'Reilly Media). As the Director of Analytics, Dan is at the forefront of Clarity's mission to providing fully integrated analytic solutions to Democratic campaigns and progressive non-profits.
Data and Democracy, edited by Andrew Therriault, former Director of Data Science for the Democratic National Committee, features eight curated essays from experts on both sides of the aisle that focus on the use of data science in the 2016 United States elections. Dan's essay, entitled, "Essentials of Modeling and Microtargeting," describes the differences between modeling and microtargeting, the basic strategy for implementation that drives their use in modern campaigns, and a few common examples of the modeling techniques developed at Clarity.
In his essay, Dan makes it clear that models, that often use polling data from as many 20,000 individual voter responses, are most effective in improving the long term efficiency of campaigns, as opposed to "horse-race" election polls that might use as few as 400 responses to illustrate short-term candidate support for a closely watched race. Deciding which models to build for a particular campaign stems directly from the targeting needs of the geographic area in question. According to Dan, "it is not just important to identify whom voters support and whether they are persuadable, but also whether or not they will take their support to the polls." To help illustrate this argument, he created an illustration that groups populations of voters based on their candidate preference and likelihood of voting. "This strategic thinking explains why turnout and support scores are the most common (and in many cases, the most useful) models for campaigns."
To learn more about the Clarity products featured in Dan's essay, and for a full list of models available at our firm, visit our catalog at http://www.claritycampaigns.com/services.