Community Spring Cleaning week is here! Join your fellow Maveryx in digging through your old posts and marking comments on them as solved. Learn more here!

Analytics

News, events, thought leadership and more.
DrDan
Alteryx Alumni (Retired)

Part II of a three part series, continue enjoying the series by reading part III

 

In Part 1 of this series, I compared the differences in a number of demographic, socioeconomic, and religious participation factors between the most Republican and Democratic leaning counties in the United States. That descriptive analysis suggested that the "typical" profiles between very Republican and very Democratic leaning counties strongly differed between these groups based on how urban they are, the area of the country where they are located, the racial and ethnic makeup of their populations, and the share of religious congregation members in different faiths. Interestingly, differences in income levels (measured on a per capita basis) were not systematically different between the two groups. While that analysis illustrates some of the characteristics that differentiate the most Republican from the most Democratic leaning counties in the country, it does not provide the underlying motivations as to why those characteristics matter, which is a topic we address here.

 


"...in times when the economy is "okay" or better, then non-economic issues, such as those embodied in the culture wars that have been a feature of American politics since the 1970s, play a more prominent role in voting behavior."


 

Early conceptual work on voting behavior was based on the hypothesis that individual economic self-interest was the major determinant, which is consistent with the notion that voters "vote their pocketbooks". While at times voters do in fact vote their pocketbooks, particularly when an election is held in the midst of an economic downturn (which helped lead to Jimmy Carter's and George H. W. Bush's failure to win second terms), in times when the economy is "okay" or better, then non-economic issues, such as those embodied in the culture wars that have been a feature of American politics since the 1970s, play a more prominent role in voting behavior.

 

The political psychologist David O. Sears and his colleagues label these factors "symbolic predispositions", but they could alternatively be described as moral and political beliefs which evoke strong emotional responses, but are not directly tied to an individual's material well-being. Examples of these relevant to the upcoming election include the pro-life versus pro-choice debate, the appropriate interpretation of the second amendment of the Constitution, and the veracity of current climate science. There is a large body of research which indicates individual voting behavior is much more closely tied to the symbolic predispositions they hold than self-interest motivations.

 


"As a result, region, race, and religion can be seen as being very strongly correlated with political beliefs, and thus partisanship, and could be referred to as the three Rs of American partisanship."


 

Symbolic predispositions are formed fairly early in life and tend to be stable. The formation of an individual's political beliefs is strongly influenced by the most common political beliefs in their community, both secular and religious. Secular communities have both a geographic component as well as a racial and ethnic component. As a result, two individuals who are from the same geographic area and come from the same racial/ethnic background and religious community are likely to hold a more similar set of symbolic predispositions than two individuals that differ on these three factors. As a result, region, race, and religion can be seen as being very strongly correlated with political beliefs, and thus partisanship, and could be referred to as the three Rs of American partisanship.

 

Another factor that is apparent in the comparison of the most Republican and Democratic counties is the impact of being in an urban or rural area (identified using population density). This factor does not fit the symbolic predisposition explanation as nicely as the three Rs, but rural areas tend to have less diversity in the set of political beliefs held within the local community, resulting in a higher probable level of commonality in the set of symbolic predispositions held by members of that community.

 

Tax, income redistribution, and social welfare policies are areas in which the two major political parties have very different platforms, and which can be most directly linked to economic self-interest. Moreover, an individual's level of income and wealth are closely tied to the potential costs and/or benefits of these policies to them. As a result, if economic self-interest is an important determinant of partisan preferences, then higher income counties should be expected to lean more Republican, while lower income counties should lean more Democratic.

 


"In the final post of this series, I will present a predictive model that examines the drivers of partisanship more rigorously than the descriptive tables provided in the first post of this series."


 

In the final post of this series, I will present a predictive model that examines the drivers of partisanship more rigorously than the descriptive tables provided in the first post of this series. By doing this, we will be able to assess the relative importance of different factors, and examine other potential drivers whose effects may be more subtle than can be readily seen using simple descriptive statistics.

Dan Putler
Chief Scientist

Dr. Dan Putler is the Chief Scientist at Alteryx, where he is responsible for developing and implementing the product road map for predictive analytics. He has over 30 years of experience in developing predictive analytics models for companies and organizations that cover a large number of industry verticals, ranging from the performing arts to B2B financial services. He is co-author of the book, “Customer and Business Analytics: Applied Data Mining for Business Decision Making Using R”, which is published by Chapman and Hall/CRC Press. Prior to joining Alteryx, Dan was a professor of marketing and marketing research at the University of British Columbia's Sauder School of Business and Purdue University’s Krannert School of Management.

Dr. Dan Putler is the Chief Scientist at Alteryx, where he is responsible for developing and implementing the product road map for predictive analytics. He has over 30 years of experience in developing predictive analytics models for companies and organizations that cover a large number of industry verticals, ranging from the performing arts to B2B financial services. He is co-author of the book, “Customer and Business Analytics: Applied Data Mining for Business Decision Making Using R”, which is published by Chapman and Hall/CRC Press. Prior to joining Alteryx, Dan was a professor of marketing and marketing research at the University of British Columbia's Sauder School of Business and Purdue University’s Krannert School of Management.

Comments
dataMack
12 - Quasar

Looking forward to the final article where you'll present the model.  I'm hoping it's in the form of an Alteryx workflow that we'll be able to tinker with to get a better understanding of your approach.