Social Context and Revenue Measures

点击翻译按钮获取中文摘要

In this project, new computational methods are applied to the texts of more than 3,000 local revenue measures that were subject to voter referendums in more and less diverse places. These data are used to investigate how tax policy characteristics may moderate the known effects of diversity on voters' willingness to pay for public spending. The results will yield new insights into what makes for a sustainable revenue source. This is crucial for enabling American society to meet the fiscal challenges of the future, and thus for the health, prosperity and welfare of American society as well as for national security. The project involves the assembly and digitization of revenue measures presented to local voters since 1986. Supervised and unsupervised machine learning techniques then are applied to classify these measures and to identify features of the text that mediate the observed associations between diversity and reported election outcomes. The most general goal of the project is to identify institutional mechanisms that moderate associations between social context and public spending. A more specific goal is to identify the causes of the well-known negative association between diversity and public spending on health, welfare, and other social transfers and services. The project thereby advances theoretical understanding of the dimensions of policy that affect voting behavior. It also analyzes institutional mechanisms that yield one of the most important observed patterns in fiscal sociology.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.