Yusuf Ziya Özcan, "Determinants of Political Behavior in Istanbul, Turkey," Party Politics, 6 (October 2000), 505-518.

First Paragraph:
Debate on political behavior in developing countries is often dominated by western concepts. This can certainly be seen in the literature on Turkey. The question is, how appropriate or valid are such approaches? This paper tackles the question by looking at the case of Istanbul. By investigating the factors that determined party preferences in the 1987 and 1991 general elections and the 1989 local elections in the metropolitan city of Istanbul, it argues that party preference is essentially an individual political behavior, which is shaped largely by the initial political socialization in the immediate environment of individuals. Reading of the western literature on voting behavior indicate that factors determining preferences and their weights vary from election to election, and identification of certain sets of factors that determine party preferences cross-culturally and over time is almost impossible. Despite the similarities in economic and social development paths, theories developed for industrialized western countries are of limited use when applied to developing countries, which are plagued with unique problems alien to the industrialized countries.

Figures and Tables:
Figure 1: Party preferences in Turkish elections since 1950 (right-wing votes combined with Islamic)
Table 1: Frequency distribution of political attitudes and preferences in Istanbul (%)
Table 2: Factors affecting party preferences in 1987 Turkish general election
Table 3: Factors affecting party preferences in 1989 local election
Table A1: Factors affecting party preferences in the 1991 general election

Last Paragraph:
The implication of these considerations is clear: any theory that claims to explain voting behavior in developing countries should accommodate factors that determine the initial formation of political preferences, such as affiliation of the family and dominant political tendencies in the area of residence. They should also be less stringent about assuming rationality in voting preferences. The consistent appearance of the earlier preference in all models developed for the three elections for Istanbul certainly confirm the need to modify the existing theories on voting.