Version classiqueVersion mobile
OpenEdition Books

Hommes armés, femmes aguerries

Fenneke Reysoo

Dispositifs d'alerte et signes précurseurs

Gender and Early Warning

Género y alerta temprana de conflictos

Début et fin des conflits

Susanne Schmeidl


This paper argues that the introduction of gender into early warning will lead to more practical, realistic and usable early warning approaches, especially if early warning is understood as a flexible system that is sensitive to the diverse situations or on the ground necessitating customised solution. A gender-sensitive approach can enhance early warning models in their basic assumption (what we consider as important or not and the questions we are asking), in their modelling (incorporating gender-sensitive indicators), and in the implementation as well (larger variety of responses options). In addition to presenting the arguments as to why a gender perspective would improve early warning in general and early warning models in particular, the paper presents suggestions for gender-sensitive indicators.

El presente trabajo plantea que la introducción de género en la alerta temprana de conflictos conducirá a un enfoque de alerta temprana más práctico, realista y útil, particularmente si la alerta temprana es entendida como un sistema flexible, sensible a las diversas situaciones o en base a la necesidad de una solución personalizada. Un enfoque sensible a la dimensión de género, puede reforzar los modelos de alerta temprana respecto a sus aspectos básicos (qué considerar como importante o no y las preguntas que se hace), en cuanto al modelo mismo (incorporando indicadores con perspectiva de género) y así como en su aplicación e implementación (ampliando la gama de respuestas posibles). Además de presentar argumentos sobre la manera de cómo un enfoque de género podría mejorar las prevenciones tempranas en general y los modelos de alerta temprana en particular, este artículo presenta también algunas propuestas de indicadores con perspectivas de género.

Cet article soutient que l’introduction du genre dans les conflits naissants amènera à des approches préventives plus réelles et plus utiles, particulièrement si les signes d’alerte sont compris comme un système flexible sensible aux diverses situations sur le terrain demandant des solutions sur mesure. Une approche sensible aux questions de genre peut mettre en valeur les modèles de préconflits dans leurs suppositions de base (ce que nous considérons comme important ou non, ainsi que les questions que nous nous posons) et dans leur façonnage (incluant des indicateurs sensibles aux questions de genre), aussi bien que dans leur mise en œuvre (possibilité d’une plus large variété de réponses). Outre les arguments établissant pourquoi l’approche genre peut améliorer le « système d’alerte anticipée » en général et les modèles de préconflits en particulier, cet article suggère aussi des propositions d’indicateurs sensibles au genre.

Texte intégral

Early warning

1Early warning is not only about assessing the possibility of conflict but also identifying the possible resurgence of conflict and the opportunities for peace (FEWER 1999, p. 3).

2The major point in early warning information collecting and analysis is not the information and analysis of the crisis in itself, but the use of this information and analysis to gain the trust of decision makers and to bring them effective response options (Adelman 1997, p. 6).

Different types of early warning indicators

31.Systemic causes – those referring to general underlying structural and deep-rooted background conditions, e.g. infant mortality, socio-political conditions, economic development.

42. Proximate causes – those relating to specific situational circumstances, e.g. widening income disparity, competition between ethnic groups, popular discontent.

53. Immediate catalysts – idiosyncratic contingent triggers, e.g. electoral fraud, purges against certain sectors of society, sharp price rises.

Proactive “response”, conflict prevention and gender considerations

6 It is suggested that gender mainstreaming would further promote appropriate action. ln other words, if policy makers and implementing agencies are aware of the differential impact of conflict on women and men, and understand that women are at risk of rape, military prostitution, economic violence and displacement in war (Tickner 1999 and 1992) the humanitarian response options they propose are more likely to be directed equally (gender sensitively) at both women and men.

7 Furthermore, if gender sensitivity exists within policy making and response institutions there is a greater chance that responding agencies will look for, and draw upon local prevention initiatives developed by women.

8 Better gender sensitivity in policy and decision-making could also have an impact on the response processes. Women in such positions could also influence the mandate, structures and processes of warning and response agencies. This must not reinforce the stereotypical image of women as being more inherently peaceful, but merely suggests that women do things differently.

9 Stereotypical images of men as strong, aggressive and rational beings, compared to women as emotional, passive and peaceful is detrimental to policy-making processes and to the development of effective and proactive response options. This “gender ideology” undermines women’s leadership as it suggests that they lack agency. It also affects men’s leadership as it inhibits their possible search for cooperative solutions. Moreover, it delegitimizes peace and cooperative strategies that may be regarded as weak alternatives or idealist. “As a concept, peace will remain a “soft” issue, utopian and unrealistic, as long as it is associated with femininity and passivity” (Tickner 1999, p. 4).

Stuctural/systemic causes

1. Political equality

10 Caprioli’s research suggests that states with lower percentages of women in parliament are more likely to use military violence to settle disputes. Only few states do currently reach above the 30 percent threshold which is argued to create a “critical mass” in terms of policy impact, which may suggest that this variable may become more powerful as women are increasingly included in political positions.

11 Female suffrage was also a significant predictor of state bellicosity. “Given two states, for example, the state having twice the number of years of female suffrage will be nearly five (4.94) times as likely to resolve international disputes without military violence” (Caprioli 2000, p. 61).

2. Economic equality

12 The percentage of women in the labor force showed statistical significance in explaining state bellicosity. This can also be construed as suggesting that women’s participation in the labor force captures the degree to which women are tied into the public sphere and other forms of participation such as voting or political activism.

3. Social equality

13 An overall measure of female fertility rate, suggests that “women’s access to reproductive health services is constrained by their broader social deprivation, including the lack of resources for, or priority to their health, restrictions on social participation, and limited access to information due to illiteracy” (p. 57; see also Sadik 1997). “Those states with high fertility rates, which represent a low status for women, were more likely to use force in international disputes” (Caprioli 2000, p. 63).

14 Marshall and Ramsey’s research (1999) using the composite measure of Gender Empowerment concurs with the above. There “appears to be a very real and robust relationship between the quality of gender empowerment and the unwillingness to use force” (Marshall and Ramsey 1999, p. 26).

15 Other research in related areas (Brumfield 1994, Erchak and Rosenfeld 1994, Levinson 1989 as cited in Caprioli 2000) suggests that “societies with high levels of family violence are more likely to rely on violent conflict resolution and are more likely to be involved in wars compared to societies with lower levels of family violence” (Caprioli 2000, p. 55). The rising level of domestic violence (or general tension) was observed in Ethiopia prior to the outbreak of war with Eritrea (Ludi 1999).

Proximate causes

16As suggested previously, minority groups, women and children are often the first affected and targeted in this way.

17• Indicators based on violations against women could contribute to improved and “earlier” early warning. For example, the recession in the status of women, discrimination against women, media scapegoating, violations of women’s human rights, and virulent attacks on women may be direct precursors of further repression and violent conflict.

18The fact that women’s human rights may be violated earlier can be connected to the fact that violence against women is often considered as a “private” and not “public” affair, and thus not related to political developments – even if performed systematically. This very subtlety makes the violation of women’s human rights (and also that of minorities) a very good indicator for “early” warning.

19Abrupt changes in women’s roles in society can also foreshadow human rights abuses and help gauge the level of tolerance in society in general. Increasing restraints on women’s roles in Kyrgyzstan, and the decline in women’s political participation throughout the former Soviet system for example is a useful indicator (Schmeidl 1999). The imposition or implementation of strict religious laws – particularly those that affect women more directly – is a useful indicator as well. For example in Pakistan the tighter imposition of Sharia laws, giving women far less rights, was a precursor to the overall strengthening of laws and a shift towards a more general conservatism in society.

20• Recent studies suggest that youth bulk and particularly a large percentage of unemployed young men be an indicator of potential instability in a society as they form the recruitment ground for militant groups.

21• The media can also play a critical role at this stage. Addressing gender related issues, the portrayal of women in the media can be a key indicator. In Rwanda, as early as 1990, extremist Hutu press and radio, began to target Tutsi women as seductresses (“femmes libres”), spies, and a threat to Hutu society (see Chretien 1995; Human Rights Watch 1996).

22Sale of valuables, particularly gold: In societies where women’s own security and wealth is based on their possession of gold and other precious materials, the sale of such items is an important indication of the level of desperation and need. Although no systematic studies have been done about women’s economic activities in the run-up to conflict, this area is worth further exploration.

Immediate catalysts

Example: sex-specific refugee migration

23Gender-disaggregated examination of flows of refugees may help prepare better responses (given that women and children represent approximately 75-80% of camp populations, Schmeidl 1997, 1999) and make better prediction as well. Linking gender and age, we can distinguish three basic waves of refugee movements (Ferris 1987). The first refugees are political exiles, or anticipatory refugees (Kunz 1973). They are politically active and flee at an early point in time foreseeing political turbulence. The second wave consists of urban dwellers, which flee after the initial outbreak of violence. The third wave of refugees (peasants), are a mass movement, composed of people who are caught in the “cross-fire” of rapidly unfolding political violence in their country and therefore flee much later in the conflict. They are the victims of violence with little involvement in the initial conflict. While the first two types of refugees tend to be educated middle-class males, the third tend to be non-educated women and children (Ferris 1987; see also d’Adesky 1991).

24In a different situation it was observed that between March and April 1992 women (especially of Muslim origin) left Priador (Bosnia and Herzegovina) by busloads. This was actually 6-8 weeks before the situation deteriorated. Muslims described this as protecting their families from danger, Serbs presented it as clearing the decks so they could prepare for all-out conflict (Last 1997).

25While it may not be fully clear if one has to look for all male (Ferris) or all female (last) migration, sex-disaggregated analysis of the fleeing groups can provide a more accurate picture of the state of events and be an effective early warning signal to conflict escalation.

Gender and early response

26Involvement of women, or gender mainstreaming

  • may lead to policy prescriptions aimed at less militarized responses,

  • can increase the possibility of obtaining gender-sensitive information giving more appropriate responses to groups at risk, and thus

  • may help develop more sensitive and comprehensive approaches to conflict early warning and prevention.

Early and non-military responses

27• It may be an overgeneralization to arguing that a heavily militarized focus, or the use of military personnel to perform non-military tasks, is linked to an extremely male-dominated decision-making process, but it cannot be denied that military operations are overwhelmingly performed by men. Even in situations where women have been active as fighters or within the ranks of the military, the overwhelming ethos has been male dominated.

28• There is some evidence that the “inclusion of women in the armed forces… might result in an alteration of the armed forces.” This supposition is supported by evidence indicating that the police force has become less violent with an increased number of female officers integrated at all levels, rather than having female officers become more violent (see Carson (1993) in Caprioli 2000, p. 54).

29• Evidence from public opinion and voting behavior also suggests that women are less likely to support militaristic response options than men – though the gap is small (Mueller 1973; Lamare 1989; Fite, Genest, and Wilcox 1990).

30• Apart from an exclusive gender-differences to gender-perspective it has been suggested that “individuals who are more supportive of equality between women and men are also less supportive of violence to resolve conflict. … In other words, societies that have internalized values that are less gender-based and apply them to their interstate and interpersonal relations should exhibit less militaristic international behavior” (Caprioli 2000, p. 55). This implies the possibility that all people can be educated toward violent or nonviolent approaches and responses to, conflict.

31Ultimately however, there is still insufficient data to prove if and how increased numbers of women in relevant decision-making positions would alter the responses of multi-lateral organizations or military operations.

Gender-sensitive information leads to more appropriate responses of groups at risk

321. Distribution of food: Research has showed that by arbitrarily assigning men to certain tasks considered as important, refugee camps often distort existing systems of division of labor among men and women (Callamard 1993). For example food production is often considered a reproductive task and thus falls under the women’s realm. But in a refugee camp the task of redistributing food is generally performed by men. Thus, relief workers (also largely male) tend to assign this work to men. The effect is two-fold. First, the supplies are often siphoned off and used in exchange for weapons or profit, and second, women’s subordination to men increases (Motsisi 1993).

332. Shelter: Another example shows that those men in charge of redistributing plastic sheeting to make huts had sold most of them for profit, while women were concerned about protecting their families during the rainy season. “The implementing partner said more sheeting was coming in. But had they changed the distribution system? No” (Marshall 1995, p. 4).

343. Provision of basic necessities: Other minor changes in planning can also have a major positive impact on assistance provided. The distribution of sanitary napkins is an obvious example. While this distribution to women refugees may seem natural, its poor implementation, by male staff members, is not. Consequently a quarter of some refugee women’s lives are sometimes wasted as they are denied access to cloth used for sanitary napkins, and are thus forced into spending a week per month in their shelters, unable to take their children to the clinic or gather firewood (Marshall 1995).

354. Camp layout: Similarly, the physical layout of a refugee camp can also play an important role in the security of refugee women. For example, many of the latrines are built side-by-side for both women and men. But women prefer risking attack in the forest rather than use these closely confined latrines. In other instances, women are forced to forage outside the camp for firewood with no protection from bandits and the militia, placing themselves in situations of possible rape and assault (Marshall 1995).

365. Limiting Sexual Violence: A 1993 evaluation of the Australian “Women at Risks” scheme concluded that the scheme’s low success rate was due to the “acceptance of the rape and sexual abuse of women as valid claim for Refugee Status” (Pittaway 1993). However, as the program was run mainly by men, the issue of rape was minimized as it became apparent that to a large extent they shared the opinion that it is “only rape” and thus not a major issue. Pittaway (1993) further believes that part of the problem, is the idea that “good men” are needed to protect women from “bad men.” This portrays women as weak and pathetic, and unable to protect themselves.

Conclusion and recommendations

371. Better Early warning: gender specificity and gender sensitivity in the construction of early warning indicators can improve medium and long-term, and in some cases, early prediction by encouraging analysts and decision-makers to pay attention to critical on-the-ground realities and details that might otherwise be missed.

382. Better responses: gender specificity and gender sensitivity when formulating responses to early warning, leads to a consideration of gender-sensitive information that can result in more realistic and sensible response options, and perhaps less militaristic/violent ones.

39More attention is needed in the following areas:

  1. Research: it is necessary to further test the relationship of gender and conflict and gender and violence – in different cultural contexts – so that early warning models can reflect on-the-ground realities and focus on long-term warning.

  2. Testing of indicators: the search for (and testing of) gender-specific early warning indicators is crucial. Most importantly, we need to test whether they indeed improve early warning models by providing contextual sensitivity and “early” warning.

  3. Analysis: the impact of gender mainstreaming on response option must be further examined. While it is argued that a gender perspective encourages the diversity of possible response options at different stages of the conflict, more exploration is needed to identify the kind of options and perspectives favored by women, how they differ from male preferences, and if/how these options would affect early warning/response processes.

  4. Support to existing systems: Comprehensive and more flexible early warning approaches (see FEWER, FAST) that can incorporate new strategies should be supported in an effort to test the impact of gender mainstreaming.

  5. Strengthen training programs: More resources and attention should be given to training programs in international organizations/governments that work on early warning and early response, so as to introduce/strengthen a gender focus.

  6. Increasing gender balance: To speed up gender mainstreaming and the inclusion of women’s perspectives in conflict early warning/response processes, programs that encourage the recruitment of women into crucial decision-making positions in international organizations/governments should be supported.


Fondation suisse pour la paix, Berne, Suisse.


Open access