Off-campus UNL users: To download campus access dissertations, please use the following link to log into our proxy server with your NU ID and password. When you are done browsing please remember to return to this page and log out.
Non-UNL users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Leadership and Economic Growth: A Text Analytics Approach
Conventional wisdom among the general public is that political leadership matters for the economy. However, economists have been slow in providing empirical evidence for such claim. The little empirical evidence that exists falls short in explaining the mechanism by which leaders affect the economy. This dissertation proposes a text analytics approach to studying the role of political leadership in economic performance, and suggests a pathway to understanding the mechanism by which political leaders may affect the economy. It posits that good leaders are committed to economic performance and that this commitment will be evident in their public announcements. Topic modeling is employed to quantify the priorities of U.S governors, as expressed in their State of the State Addresses. The estimated priorities are then used to construct a consistency measure. We find strong evidence that consistency on priorities predicts measures of economic performance. Using classical Canonical Correlation Analysis (CCA), we validate the approach by showing that the thematic contents in the speeches mirror objective measures of actual future state budgets. The approach developed and expounded upon in this dissertation shows that a leader’s commitment to economic performance can be measured objectively and that this commitment has real and measurable consequences. After motivating the research goals of this dissertation in Chapter 1, Chapter 2 offers an extensive survey of topic modeling, from a multivariate regression perspective to correlated topic modeling. Chapter 3 reviews the Bayesian treatment of CCA with the aim of providing the building blocks of Chapter 5. Chapter 4 applies Latent Dirichlet Allocation, Bayesian linear regression, and classical CCA to show that leaders' priorities can be measured, and that consistency over priorities predicts economic performance. Chapter 5 builds on the previous chapters to propose a supervised topic modeling algorithm that simultaneously estimates the topics within documents and the association between the themes and economic outcome variables. Chapter 6 summarizes this dissertation.
Bikienga, Salfo, "Leadership and Economic Growth: A Text Analytics Approach" (2018). ETD collection for University of Nebraska - Lincoln. AAI13420167.