In this dissertation, I examine the theoretical and methodological bases for drawing inferences about Americans’ fiscal policy preferences. American fiscal policy preferences are somewhat contradictory, with Americans expressing high levels of support for helping the poor, but low levels of support for fiscal policies like welfare that help the poor. Existing scholarship attributes this to individual-level racism, a set of beliefs that link race to merit and deservingness of aid, and a norm of equal treatment. However, I argue past survey-based research is limited by the topics assessed and compatibility of item wording across surveys. To broaden the fiscal policies assessed, I conduct three primary survey experiments that compare preferences for low-risk, particularistic fiscal policies (e.g., Welfare) to high-risk, universalistic fiscal policies (e.g., Universal Basic Income). I find support for my hypotheses that American fiscal policy preferences are driven by perceptions of whether fiscal policies treat people equally, consistent with past research on the fiscal policy preferences in the United States, United Kingdom, and Germany, with Americans tending to prefer fiscal policies they perceive as treating people more equally. This effect persists after controlling for individual-level racism, demographic variables, and other factors using a psychometric model developed in this dissertation to account for multicollinearity among variables. A preference for fiscal policy universalism is consistent with the paradox of redistribution in sociology, and a preference for equal treatment is consistent with a strategic theory of social identity advanced in this dissertation. Controlling for policy risk, Americans prefer the fiscal universalism of Social Security and Universal Basic Income over other, particularistic fiscal policies like Reparations and Welfare. To facilitate objective comparisons of items across surveys, I advance a computational psycholinguistic theory of survey research that allows for quantitative textual analysis of survey items, news articles, and other texts using the deterministic, automated hyperspace analogue to language. HAL does not require human judgement, allowing researchers to control for the influence of environmental heuristics (e.g., news articles) to identify true-score values more accurately for individual level traits, like the trait preference for equal treatment I hypothesize drives American fiscal policy preferences. I conclude with an examination of the emerging science of neurodiversity—the opposite of eugenics—connecting it to republican political theory.