Graduation Year


Date of Submission


Document Type

Open Access Senior Thesis

Degree Name

Bachelor of Arts



Second Department


Reader 1

Eric Helland

Reader 2

John Pitney

Terms of Use & License Information

Terms of Use for work posted in Scholarship@Claremont.

Rights Information

© 2019 James L Dail


This thesis builds off of the work of Robert Putnam and Charles Murray by using a fixed effects regression model and a first difference regression model to see if there is a relationship between the number of community institutions within a county and that county’s resiliency against social decay. Historically, it has been difficult to determine if there is a causal relationship in either direction between these variables because they are endogenous. That is, the underlying factors that determine whether a county will have low levels of social capital also determine whether the same county will have high levels of social decay. For example, income level is strongly correlated with both variables, and even if one controls for income level, it is difficult to control for the advantages that come with being wealthy and lead to higher levels of social capital. Due to this, it is difficult to say whether a higher income level leads to more social capital and increased resiliency against decay simultaneously, or if higher social capital itself causes increased resiliency. This thesis counteracts this endogeneity by introducing an exogeneous factor into the data, the decline in manufacturing employment. Using data primarily from the Census Bureau, Pennsylvania State’s Northeast Center for Rural Development, and the China Shock Project, I find that there is little evidence for any relationship between social capital and resiliency against social decay. The one exception to this is religious organizations, for which a clear, positive effect exists across multiple variables and both regression models. Further research should focus on different approaches to measuring social capital and expand the scope of this thesis by including additional variables, as well as by using survey data about community participation instead of relying on the number of institutions in an area.