UC Berkeley

This dissertation develops a general quantitative model of field structure for analyzing the field theoretic properties of social phenomena. I use core structural principles found in the dominant variants of field theory (Lewin, Bourdieu, Dimaggio and Powell, Martin, and Fligstein and McAdam) to outline and elaborate a broad class of models consistent with the major assertions of field theory. These models use nonparametric local regressions to estimate field effect over multidimensional social space. By using univariate and multivariate statistics, these models enable determination of the existence or nonexistence of the field, the best description for the axes of the field and the complex patterns of variation in field effects over social space.

In three distinct analyses based on data from the Consumer Expenditure Survey and the Panel Study of Income Dynamics, I demonstrate how the organization of consumption as whole has changed over time and how individuals at different points in social space have been affected. Using a synthetic cohort design, I show how lifecycle consumption is differentiated by age, income, and education, and how these effects interact and moderate each other. Finally, I analyze how households changed their consumption in response the Great Recession. I find that cultural change in consumption preferences are a better explanation for changes in consumption than recession-related economic shocks or drops in consumer confidence.

This dissertation makes two contributions. First, it attempts to further the development of field theory by outlining and demonstrating a quantitative model of field structure. This model offers an improvement over the current state of field modeling in terms of statistical rigor and flexibility. The model I present here enables more detailed and careful analysis of field structure and transformation and is general enough to help adjudicate between competing claims of conflicting field theories. Second, this dissertation highlights the necessity of considering economic and cultural determinants of consumption in conjunction. Consumption patterns are highly contingent on the specific conditions under which individuals make consumption decisions. While income, wealth, and savings are important, age, education, and family size are equally and frequently more important for predicting variations in consumption.