FLINT, COLIN ROBERT; PHD

                         UNIVERSITY OF COLORADO AT BOULDER, 1995
 
                         GEOGRAPHY (0366); HISTORY, MODERN (0582); POLITICAL SCIENCE, GENERAL (0615)
 

                         This thesis is an inquiry into the spatial diffusion of the Nazi party's electoral support in the seven
                         Reichstag elections held between 1924 and 1933. Socio-economic and voting data were taken from the
                         Wahl-und Sozialdaten der Kreise und Gemeinden des Deutschen Reiches, 1920-1933 at the Central
                         Archive of the University of Cologne. The division of Germany into eight cultural-historical regions allows
                         for an examination of the regional specificity of the Nazi party's electorate. This thesis challenges
                         previous analyses which have treated the German electorate as a national unit. The spatial diffusion of
                         the Nazi party vote is modelled using structural-spatial regression models. The dependent variables are
                         the percentage change in the Nazi party vote between consecutive Reichstag elections. The modelling
                         is conducted at the national scale as well as at the scale of the eight regions. Structural-spatial models
                         include socio-economic variables as well as geographic variables to demonstrate the importance of
                         spatial and contextual effects. Two concepts are incorporated into the analysis of aggregate data, spatial
                         heterogeneity and spatial dependence. The presence of spatial dependence indicates spatial diffusion
                         of the Nazi party vote at the local scale. The mapping of the change in the Nazi party vote illustrates local
                         scale contagious diffusion and regional scale hierarchical diffusion. I demonstrate the domain-specific
                         nature of the Nazi party's electoral support which national models fail to include. Though, at the national
                         scale, the Nazi party's electorate was cross-class in nature, different sections of society were more or less
                         likely to vote for the NSDAP according to spatial and temporal context. The ability of the Nazi party to
                         extend its electoral support across space was limited both spatially and temporally. Established theories
                         which have been used to explain the electoral support of the Nazi party are shown to be context specific
                         rather than applicable to Germany as a whole.

Social Systems Simulation Group P.O. Box 6904 San Diego, CA  92166-0904 Roland Werner, Principal Phone/FAX  (619) 660-1603