Despite decades of interdisciplinary research on trust, issues of conceptualization and measurement remain, which has created a fragmented literature with little consensus about the origins of trust. To address these shortcomings, my project tackles the following questions: What is trust? How do we measure trust? And where does trust come from? I do so by offering a concrete and measurable definition of trust—what I call relational trust—as well as a research design—vignette experiments—that captures its core elements. I decompose the sources of trust into four analytical components: characteristics of the truster, characteristics of the trustee, characteristics of the exchange relationship, and characteristics of the social forces external to or beyond the exchange relationship. From this, I propose a number of novel and pre-existent hypotheses addressing each major source. My secondary goals in this dissertation are to (a) investigate the relationship between political institutions and an alternative microfoundation known as general social trust, and (b) explore if and how general social trust impacts relational trust. I argue that my new conceptualization, unique research design, and novel propositions identifies foundations of trust that have yet to be observed in the literature.
This project employs two sources of data. First, I collected data from a variety of cross-national public-opinion data sets (including the World Values Survey and the European Values Study) to form a repeated cross-sectional time-series pseudo-panel spanning 30 years (1980 to 2009), 74 countries, and 248 observations. With this data, I employed linear regressions, random-effects models, and fixed-effects models to test hypotheses regarding the relationship between political institutions and general social trust. Second, I administered two novel vignette experiments to two distinct populations of Amazon.com’s Mechanical Turk workers (N = 1,276 and N = 1,322) and University of Washington undergraduate students (N = 884 and N = 841). With this data, I employed hierarchical linear regressions and multilevel mixed-effects multiple mediator models to investigate and adjudicate between the four analytical sources of relational trust.
Taken together, the empirical chapters show that trust has varied and numerous foundations. National context in the form of political institutions explains temporal variations in general social trust within and across nations. I find that characteristics of a truster (e.g., general social trust and particular social trust), characteristics of a trustee (e.g., perceived competence and motivation), and social forces external to a truster-trustee exchange relationship (e.g., social constraints) all effect relational trust, while weak evidence is detected for characteristics of a truster-trustee exchange relationship (i.e., social identity). I also find that causal attributions and social constraints interact in their influence on relational trust, and that other-praising emotions such as gratitude and admiration serve as plausible mechanisms connecting perceived motivations to relational trust. Overall, perceived motivations and commitment (a characteristic of a trustee) exert the greatest impact on relational trust. These results support the idea that people rely on multiple sources of information within, between, and beyond individuals when forming trust.