Predicting Political Involvement through Demographics, Overall Involvement, and Political Interest

Political participation and involvement has been one of the keys to democratic success since the founding of the United States (McFarland & Thomas, 2006). In some contexts (say when disapproval of the government as a whole is at an all-time high) participation, or lack thereof, is critical in regards to prosperity. Individual citizens do not see the point in a government, and get discouraged from participating in it (Shames, 2014). A democratic government relies on its’ citizens to play some kind of role in the system-running to represent others, voicing an opinion or helping someone else become an effective political leader, in order to run effectively and stay afloat. The many questions underlying political involvement for decades have been “how do people become politically active?”, “to what extent are they politically active?” and “what are the consequences of that activism?”, while usually expressed more or less eloquently (McFarland & Thomas, 2006; Eulau & Schneider, 1956; Snell 2010; Prior, 2005, 2010). A multitude of factors can contribute to political involvement, making these questions that much harder to definitively answer- similar to the classic nature versus nurture debate. Scholars and corporations have been actively searching out answers to these questions, and their alternatives, for years and have discovered many factors can matter but that the real answer is simple- it depends (McFarland & Thomas, 2006; Jones-Correa & Leal, 2001; Smith, 2013).

Narrowing the contributing factors down to one or two holy-grail characteristics has proven to be improbable, if not impossible. The direct and observable factors that are consistently cited as influencing participation are often too far developed for any researcher to manipulate or control in a true experiment. So, the next best option is to survey the masses and hope that randomization will filter out some of these effects. But, causation can never be inferred with studies designed this way, making the process of pinpointing what causes involvement that much more challenging. For example- Pew Research Center, one of the most reputable think tanks in the world, conducted a study examining social media’s effect on civic engagement (Pew Research Center, 2013). Their question- will social media’s presence negate the traditional finding that only those with high socioeconomic status participate in politics? Their answer- Nope… Well, not directly at least. Those with higher socioeconomic status utilize social media as an outlet for political activity as much as they participate in person, and this effect held true for those at the opposite end of the scale. When access to social media is held constant, those with lower socioeconomic status simply weren’t as actively involved. Their proposed reasoning for this finding was that those with lower socioeconomic status aren’t exposed to political discussion or activity in the same way as those with higher socioeconomic status, so they just don’t have the same desire to be involved, and may not see what they can gain from such an involvement.

Many studies attempt to relate one or two kinds of influential characteristics (i.e. involvement in extracurricular activities at school) to one kind of political involvement (i.e. identifying as “political”). Pancer, Pratt, Hunsberger and Alisat (2007) identified many behaviors/traits that predicted involvement for their sample, but the ways in which those people were involved may have been too similar to each other, resulting in a narrowed sense of what “involvement” is. This analysis identified nine different traits commonly used to define political involvement:

· Having ever communicated thoughts to a public official (Pancer, Pratt, Hunsberger & Alisat, 2007; Smith, 2013; Pew, 2014)

· Having ever held a governmental office (Smith, 2013)

· Having ever worked on a political campaign (Pancer, Pratt, Hunsberger & Alisat,2007; Smith, 2013; McFarland & Thomas, 2006; Pew 2014)

· Having ever attended a political rally (Pancer, Pratt, Hunsberger & Alisat,2007; Smith, 2013; Eulau & Schneider, 1956; Pew 2014)

· Having ever contributed money to a political cause/party/candidate (Smith, 2013; Eulau & Schneider, 1956; Pew Research Center; 2014)

· Having ever sought out political information through three different channels-reading the newspaper, online or listening to the radio (Smith, 2013; Eulau & Schneider, 1956; Prior, 2005)

· Having ever discussed politics with others (Eulau & Schneider, 1956; Prior, 2005; Pew Research Center, 2014)

Utilizing others’ significant definitions in combination will allow for a more representative picture of involvement than offered in previous studies.

The characteristics that have been associated with different kinds of political involvement in previous research range from definitive characteristics such as age, to manipulative characteristics such as number of hours spent volunteering a week. The characteristics used in these analyses can be broken up into three groups, which have been supported in previous research as predictors as well: demographics, general involvement and political interest.

· Demographic characteristics: perception and preference of politics, age and gender (Shames, 2014; Pew Research Center, 2014; Prior, 2010).

· General involvement characteristics: club membership, club leadership, time allocated for clubs and regular religious service attendance (Pancer, Pratt, Hunsperger & Alisat, 2007; McFarland & Thomas, 2006; Jones-Correa & Leal, 2001; Pew Research Center, 2014; Jansen, 2001; Prior, 2010).

· Political interest characteristics: political fascination, strength of political feelings, number of days per week they read the newspaper/ watch television/ listen to the radio/ read on the internet political information, identifying as a partisan, or identifying as someone interested in politics (Smith, 2013; Pancer, Pratt, Hunsperger & Alisat, 2007; Eulau & Schneider, 1956; McFarland & Thomas, 2006; Shames, 2014; Prior, 2005; Pew Research Center, 2014; Rampell, 2014; Prior, 2010).

All of the factors in this analysis have significantly predicted some degree of political involvement in previous research, but have not been observed together as groups of characteristics predicting different kinds of involvement.

Hypothesis one: Demographic characteristics alone should not be able to predict political involvement as well as the full model also including the general involvement model and the political interest model. Majority of the previous research done concludes similarly- demographics matter, but other factors contribute to involvement more and are therefore more important. However there are significant implications if this analysis finds that demographic variables can predict involvement just as well as this all other kinds of predictors when kinds of involvement are compiled.

Hypothesis two: The general involvement model should be able to predict political involvement as well as the full model that includes the demographics model and the political interest model. Previous research has found that being involved in other aspects of life is related to being involved politically. But, the nature vs nurture debate can come into play again here. If this analysis yields consistent results, then kinds of general involvement might be a better way to look at kinds of political involvement.

Hypothesis three: The political interest model should be able to predict political involvement as well as the full model that includes the demographics model and the general involvement model. Previous research has found that interest alone is a strong predictor of involvement, but interest alone can’t instantly catalyze all kinds of involvement, and other predictors are needed to capture the full involvement picture.

Hypothesis four: The political interest model should be able to predict political interest better than the other two reduced models, and will have the strongest single predictor of involvement in it. Since demographics and general involvement have traditionally predicted political interest, the compounding relationship should show up in this analysis as well.