Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)

First Advisor

David J. Smyth


This thesis examines several issues dealing with the public's social preference function between inflation and unemployment. First, the preference function is estimated as a nonlinear function. Second, the thesis analyzes the ability of a presidential administration to improve its popularity level by exploiting a short-run Phillips curve. Finally, preference functions disaggregated by sex, race, political party, and geographical region are estimated. Presidential approval levels are proxied by monthly Gallup Poll approval data. Chapter One serves as an overview of the current body of literature. While theoretical analyses of macroeconomic policy have traditionally assumed that the public's social preference function is concave to the origin, in practice most empirical studies estimate a linear relationship between presidential popularity, inflation, and unemployment. The majority of the empirical work assumes that the impact of macroeconomic variables on political popularity remains constant across presidential regimes. Chapter Two applies appropriate tests for structural change to the data. The results indicate that structural change in economic perceptions does take place between presidential regimes and thus that each regime should be estimated separately. The econometric study in Chapter Three examines the popularity of each president from Eisenhower to Reagan. For a majority of the presidential administrations, it demonstrates that the public's trade-off between inflation and unemployment is nonlinear. This nonlinearity indicates that the marginal rate of substitution between inflation and unemployment varies with the relative rates of the macroeconomic variables. A major question of policy importance is whether or not an administration can increase its popularity in the short-run by exploiting a short-run Phillips curve. Chapter Four estimates the Phillips curves and provides evidence that while it is theoretically possible for a president to create a political business cycle, the gains in popularity are so small as to make the activity not worthwhile. Chapter Five examines presidential approval on a disaggregated level. There exists extensive Gallup Poll data disaggregated by sex, race, political party, and geographical region. Nonlinear social preference functions are estimated for each of these groups. Chapter Six provides a summary of the results discussed in the thesis.