A statistical analysis of Oregon's land-use ballot initiative

Abstract

This paper explores predictors of support for land-use ballot initiatives in Oregon. Possible predictors include a variety of socioeconomic, spatial, and political variables. Multiple least squares regressions find that aggregate levels of education within a county and the proportion of a county's workforce engaged in farming or forestry occupations appear to play a strong role in county-level land-use ballot initiative results. These effects of education and industry sector employment are consistent across multiple elections. Real per capita income and party identification appear to play a role in specific elections and are not consistent predictors. Analysis from initiatives in 1970, 1976, 1978, 1982, 2000, and 2004 from Oregon are used to draw these conclusions.