Women in uniform: the opening of combat roles in state militaries

ABSTRACT

Women have historically been excluded from combat roles in state militaries. However, in recent years, women’s growing involvement in combat roles has sparked public debate. Currently, only a small minority of countries allow women into their ground combat forces. Given the policy relevance, it is important to examine the conditions under which militaries will allow women into combat roles. Using data from 1970–2016, I empirically examine how a variety of aspects of women’s participation in social, political, and economic activities and institutions affects the probability that combat roles will be opened to them. The results provide robust evidence that women’s participation in politics, civil society, and economic activities are associated with a higher probability that these exclusionary policies are lifted from combat roles. However, I do not find evidence of an association between expectations pertaining to women’s familial roles and the probability of combat positions being opened to them.

Históricamente, se ha excluido a las mujeres de las funciones de combate en los ejércitos estatales. Sin embargo, en los últimos años, la creciente participación de las mujeres en las funciones de combate ha generado un debate póblico. En la actualidad, solo en una pequeña minoría de países se permite que las mujeres formen parte de las fuerzas de combate terrestre. Dada la relevancia política, es importante analizar las condiciones en las que los ejércitos permiten que las mujeres desempeñen funciones de combate. Mediante el empleo de datos de 1970 a 2016, analizo empíricamente la forma en que diversos aspectos de la participación de las mujeres en instituciones y actividades de índole social, político y económico influyen en la probabilidad de que puedan desempeñar funciones de combate. Los resultados proporcionan pruebas contundentes de que la participación de las mujeres en la política, la sociedad civil y las actividades económicas se asocia con una mayor probabilidad de que dichas políticas de exclusión se eliminen de las funciones de combate. Sin embargo, no encuentro pruebas de una asociación entre las expectativas relativas a los roles familiares de las mujeres y la probabilidad de que ocupen posiciones de combate.

Les femmes ont historiquement été exclues des rôles de combat dans les armées d’État. Cependant, ces dernières années, l’implication croissante des femmes dans des rôles de combat a déclenché un débat public. Actuellement, seule une petite minorité de pays autorisent les femmes dans leurs forces de combat au sol. Dans un souci de pertinence politique, il est important d’examiner les conditions dans lesquelles les armées autoriseront les femmes à endosser des rôles de combat. J’ai utilisé des données datant de 1970 à 2016 pour étudier empiriquement la manière dont divers aspects de la participation des femmes aux activités et institutions sociales, politiques et économiques affectent la probabilité que des rôles de combat leur soient rendus accessibles. Mes conclusions fournissent des preuves solides que la participation des femmes aux activités politiques, économiques et de la société civile est associée à une plus grande probabilité que ces politiques d’exclusion affectant les rôles de combat soient abandonnées. Je ne parviens toutefois pas à trouver des preuves d’une association entre les attentes relatives aux rôles familiaux des femmes et la probabilité que des postes de combat leur soient accessibles.

KEYWORDS:

• Women and the military
• gender and conflict
• international security
• civil-military relations
• survival analysis

Supplementary Material

Supplemental data for this article can be accessed on the publisher’s website.

Notes

¹ There is a growing body of literature that employs quantitative methods to study women’s participation in a variety of kinds of political violence. This includes women’s participation in non-state armed groups (Dalton and Asal 2011; Henshaw 2016a, 2016b; Thomas and Bond 2015; Thomas and Wood 2018; Wood and Thomas 2017); female peacekeepers (Karim and Beardsley 2013); and internal security sectors of countries (Huber and Karim 2018).

² Obradovic’s analysis differs from Carreiras’ in that it includes the newer NATO member states in Eastern Europe and that it tests several more correlations.

³ However, Pindyck (2019) is currently developing a comprehensive dataset that focuses on women’s integration into state militaries, in a variety of dimensions, cross-nationally and over time.

⁴ There are many parallels between women’s participation in non-state armed groups and state militaries, however, there are also key differences. While upwards of 40% of rebel groups employ women in combat roles (Wood et al. 2017), only a handful of state militaries do. Part of the discrepancy in these recruitment patterns might be driven by differences in external threats. State failure is relatively rare, especially in recent history. However, most rebel groups face near constant threat of demise, and thus, are likely more willing to expand their recruitment pool.

⁵ Even when militaries have opened combat roles to women in the past – such as the Soviet Union during World War II – women’s status often did not improve following the conflict (DeGroot 2001; Enloe 1980; Moore 1991; Segal 1995; Seitz, Lobao, and Treadway 1993). However, I still expect that military leaders and policy makers will be more resistant to the possibility of expanding the roles of women in less gender egalitarian societies.

⁶ Doing so drops just under 900 observations, leaving a sample of 6,881 country-year observations.

⁷ The list of countries identified by Fitriani, Cooper, and Matthews is displayed in Appendix Table 1 and the summary statistics for all variables in Appendix Table 2.

⁸ Fitriani, Cooper, and Matthews identify the first year in which legal action began in countries to open combat roles to women. The authors also provide alternative dates for some countries (though many remain the same), based on when GCC roles were fully opened to women. The core results remain the same when these alternative dates are used.

⁹ North Korea does not enter the data because it opened combat roles to women in the 1950s, and thus, experienced the “failure” event before the start year of the analysis. However, it is included in the logistic regression models (see robustness checks).

¹⁰ An additional advantage of the peace data is that it is collected through 2015, with information on the rivalries that extend into 2016, meaning it is available for the entire time period of this analysis. The MID data, however, is available only through 2010.

¹¹ Note that coefficients are reported rather than hazard ratios.

¹² The characteristics of countries that open combat roles appear to be fairly consistent over time. The mean values of the main independent variables for countries that opened combat roles to women at first appear to be noticeably lower before 2000, suggesting that early movers had a lower threshold of gender equality. However, when Eritrea (opened GCC roles in 1998) is taken out, the mean values are much closer. For instance, before 2000 (with the exception of Eritrea) the mean level of women’s political participation for these states was 0.954 and the average fertility rate was 1.758. Post-2000, these averages were 0.963 and 1.764 respectively. The greatest difference is in women’s civil society participation (0.833 vs. 0.915). This difference is likely explained by the increase in general gender-equality legislation (Fitriani, Cooper, and Matthews 2016), which is often supported by women’s civil society organizations.

¹³ Of the 16 countries that opened combat roles to women prior to 2008 (when conscription data ends), only 2 (Canada and New Zealand), relied on voluntary recruitment. This is likely the result of some countries using conscription and women in response to existential threats (e.g., Israel and Eritrea), while others conscript men and women as part of an effort to promote gender equality (e.g., Sweden). However, as countries increasingly abolish conscription, the role of this variable will likely diminish.