Madison Dayton Daines and Quin Monson, Department of Political Science
Introduction
I believe that the religion a person was socialized in has lingering affects on his or her political values. Parents play the most prominent role in the political direction of their children both early in life and up through young adulthood (Niemi and Jennings 1991). This is because parents transmit political values to their children through socialization, the everyday conversations and interactions about the political world as well as broader norms and values that occur within the family, thus influencing their political attitudes and behavior. Religious ties also greatly contribute to individuals’ political attitudes and behaviors (Wald and Calhoun-Brown 2010). In America, religion plays a prominent role in society and thus also in political socialization. Pew’s 2007 Religious Landscape Survey found that 83% of the American population affiliates with a specific religion, 78% of those identify with a Christian religion. This research will address the relationship between religion and political socialization. Put simply: parents socialize children, and religion influences political behavior. Thus, religious socialization from parents should play a prominent role in socializing political behavior. I hypothesize that religious and political socialization interacts to influence political attitudes and behaviors.
Methodology
To carry out my research design, I analyzed Kent Jennings’ Youth-Parent Socialization Panel Study, 1965-1997. Specifically, I analyzed the data from the first wave in 1965 as well as last wave in 1997. This dataset is a panel study designed to measure political permanence and changes across time between generations. With over 150 measurements of religious denomination, this study also accurately captures religious affiliation. This data contains 1000 survey participants.
For this analysis, I used robust OLS regression models because the independent variables were continuous. The main independent variable used was an interaction variable composed of the participant’s mother’s party identification interacted with the level of religiosity of the family. Measures of childhood religiosity and mother’s party identification were used because they accurately capture the effects of socialization. This interaction allowed me to analyze the simultaneous influence of childhood religiosity and mother’s ideology instead of considering these two separate, additive factors. The dependent variable is the participant’s party identification as of 1997. Appropriate control variables were also implemented. Note that only Baptists were analyzed for each regression model, so as to control for religious denomination factors. An evangelical dummy variable was preferred, but data limitations required an alternative.
Results
Parent to Child (1965)-The first regression involved the parents’ data from the 1965 wave and their children’s data from the 1997 wave. The dependent variable was the Party ID of the child in 1997. The independent variable was an interaction between the mother’s Party ID in 1965 and the church attendance frequency of the family in 1965, contingent upon the family being Baptist. Two control variables were also used: whether or not the child became less religious since 1982 and child’s frequency of church attendance as of 1997. The independent variables were statistically significant at the .1 level. The significance of the interaction variable shows that there is an interactive link between mother’s party ID and level of religiosity growing up. Together, they predict party ID of the child with some level of significance, thus supporting my theory. The control variables in the model did not appear to be statistically significant. Because the independent variables were significant and the control variables were not, this means that the socialization of religion and level of religiosity pointedly contribute to determining Party ID, while the general effects of religiosity does not.
Parent to Child (1997)-Since the first group of participants were high school seniors in 1965, they were likely to be socialized politically in the early to mid fifties. According to culture war theory in America, conflict between values to be considered conservative and values considered liberal did not emanate until the 1960’s (Davison 1991), which lead to the prominence of the Christian Right in the 1970’s (Williams). Religions would not have been politically segregated before then because values were not in conflict. Because this information would suggest more significant results in later generations, I decided to conduct the same test for the children of the parents in 1997.
Instead of using the second generation’s children’s party ID as the dependent variable, the third generation’s children’s party ID is now the dependent variable, and the parent’s party ID, religion, and level of religiosity are the independent variables. The control variable included in the model is the child’s current frequency of church attendance. The regression provided surprising results. Parent’s party ID became extremely significant at the .01 level, but the childhood church frequency variable and the interaction variable produced the opposite result. Instead of having more significant results, the significance of the religiosity virtually disappears all together. The interaction variable’s coefficient even became negative, indicating an inverse relationship between the interaction of parent’s party ID with the level of religiosity and being Republican. The control variable, current frequency of church attendance, was not significant either. This means that neither religiosity measurement for Baptists were meaningful in predicting party ID. These results invalidate the notion that religious factors mattered at all in predicting party ID, something that numerous others have found to be true. According to these results, neither the childhood religiosity nor adulthood religiosity influenced the party ID of the Baptist participants socialized in the 1980’s.
Conclusion
My theory did not entirely hold up to statistical analysis, but the analysis did produce interesting results. Although the interaction between parent party ID and childhood religiosity was significant at the .1 level for the Baptists socialized in the 1950’s, the same interaction produced the opposite of the expected results for the Baptists socialized roughly in the 1980’s. Explanations to these results require further tests and analysis. Some possible theories to explore that would explain these results include: people are not as politically aware as they were in 1965, religion had a different meaning to people in 1965 than it does now, other factors besides religion influence party ID more now than it did in 1965, or people have become more secular since 1965.
Discussion
This research project was presented at the Mary Lou Fulton conference held at BYU as well as a CSED workshop. I received a lot of helpful feedback from these forums that have encouraged me to continue with this research. From this experience I was able conduct original research, discover new results, and broaden my understanding of political socialization in a religious context, which has been very rewarding.