Transition to first marriage in Turkey

DeAnna L. Gore, Florida State University
Elwood Carlson, Florida State University

Using data from the 1998 Turkish Demographic and Health Survey, we use STATA and AML to show how event history analysis can be used to calculate the non-proportional piecewise linear estimate of the hazard for first marriage among women in Turkey. Our analysis explores the minority group status hypothesis by examining marriage timing for Turkish and Kurdish women with respect to education. Members of a minority population within a larger society sometimes exhibit differences in key demographic processes that cannot be explained away in terms of other compositional effects, such as levels of education. An interaction of ethnicity and education across different age segments show earlier marriage for Kurdish-speaking than for Turkish-speaking women at low education levels, but later and less marriage for Kurdish women at high education levels.