Abstract
Gender inequality has been empirically linked to the incidence of terrorism, as a motivator for women’s involvement in terrorism, and political and social violence more generally. Although these studies demonstrate that advances in gender equality on average decrease terrorism, it is unclear how these influences translate to individual nations or conflicts. Because Turkey has witnessed consistent but unequal regional advances in gender equality, it provides an ideal setting to examine how these developments have influenced terrorist violence. Analyzing data from Turkey’s 81 provinces from 2000–2013, this study finds important regional differences in the influence of gender equality on terrorism.
Notes
Notes
1 Specifically, these studies examined between 57 (Salman, Citation2015) and 165 nations (Berrebi & Ostwald, Citation2016), covering a period between 9 (Salman, Citation2015) and 35 years (Harris and Milton, Citation2016). The dependent variable for each of these studies was the incidence of terrorism and/or terrorism fatalities, and these data were drawn from either the International Attributes of Terrorism database and/or the Global Terrorism Database (GTD).
2 See Wood and Ramirez (Citation2018) for a complete and detailed account of the discrepancy between the proposed theoretical mechanisms and the empirical methodologies employed by these studies.
3 For a full description of these attacks see https://www.start.umd.edu/gtd/search/Results.aspx?start_yearonly=&end_yearonly=&start_year=&start_month=&start_day=&end_year=&end_month=&end_day=&country=209&perpetrator=449&dtp2=all&success=yes&casualties_type=b&casualties_max=.
4 “The marriage of underage girls; the demand for baslik (brideprice) in the marriage contract; the denial of girls' rights to education; and the emphasis on women's fertility were continuing signs of the uneven socioeconomic development of the country” (Kandiyoti & Kandiyoti, Citation1987, p. 322).
5 This variable was calculated as the (age of grooms – age of brides) within each province within each year.
6 As the female proportion of legislators was measured at the national level, any findings specifically related to the relationship between this variable and the outcome variables should be treated with caution, as the standard errors for these estimates will be smaller than they would be if province-level variations were observed, potentially leading to Type I errors.
7 Sensitivity analyses were also conducted that excluded the number of terrorists killed from each attack. These additional analyses had no meaningful impact on the substantive findings.
8 Sensitivity analyses also revealed that these null findings were not due to multicollinearity despite their relatively high correlation with other variables.
9 See Sari (Citation2012) and Kılınç, Wyatt, and Richardson (Citation2012) for Turkey specifically; see Coffey and Delamont (Citation2000) for a broader perspective.