The scale of exclusion

(1)

The scale of exclusion

A research on humanitarian organizations ex- or including men as a result of gendered norms

Nijmegen School of Management

Master thesis Political Science: Gender, Equality, Diversity and Inclusion in Political Science

Annelotte Heutz (s4580419)

Radboud University Nijmegen

Supervisor: Dr. (Ph.D.) Jutta M. Joachim

Words: 24.685

(2)

Abstract

Humanitarian organizations claim to cover for the seven fundamental principles of providing humanitarian assistance. Back in 2003, Carpenter showed with her analysis on humanitarian organizations that gendered norms are responsible for the exclusion of men in humanitarian assistance which violate the core principles. Since this publication many things have changed. Olivius even showed with her findings on humanitarian organizations that the organizations no longer exclude men from their assistance. Instead, the organizations adopted the modern liberal values which caused the implementation of modern gender beliefs and roles. Due to this, men are also categorized as possible victims which are in need of aid. This results in men no longer being excluded from humanitarian assistance solely based on sex and age. This paper researched both arguments in order to provide an answer to the research question which posed that to what extent humanitarian organizations use gendered norms? Researching the humanitarian organizations INTERSOS and Doctors Without Borders, by carrying out a content analysis on their general webpages and the webpages on the humanitarian crisis in Yemen, showed that there is a scale of excluding men in humanitarian organizations and assistance. On the one hand humanitarian organizations, based on the findings on INTERSOS, which still predominantly recognize women and children and exclude men. On the other hand organizations, based on the findings on DWB, including and recognizing men almost completely. The only step these organizations have to take in order to fully account for men is also recognizing men in policy and not only throughout imagery.

Key words: excluding men, humanitarian organization, gendered norms, traditional gender roles, modern gender roles, Yemen, INTEROS, Doctors Without Borders.

(3)

If someone opens the website of the UN Refugee Agency, an image of a young mother with her newborn baby appears (UNCHR, 2019). If then someone goes to the page in which the organization explains the protection it provides, an image of three young children will be shown. The same holds for the website of UNICEF (2019), picturing four young children and again a mother with her newborn baby. With the well-known saying that an image speaks louder than words, it seems that both humanitarian organizations follow what Carpenter (2003) refers to as the women and children first rule which is a typical example of gendered norms in organizations. These gendered norms are derived from gender beliefs, which Carpenter (2003) showed in her analysis on humanitarian organizations. Gendered norms and beliefs are responsible for the exclusion of men as victims in need of humanitarian assistance (Carpenter, 2003). However, Carpenter (2003) did her analysis back in 2003 and ever since many things regarding gender beliefs and gendered norms have been changed. Scholars even showed that Carpenter's argument of the exclusion of men in humanitarian organizations no longer applies to the case of today. Olivius (2016) even argued that humanitarian organizations nowadays have fully adopted the modern liberal values. These values are responsible for the organization using the modernization approach in both their organization as well as their interventions, which cause modern gender roles. These modern gender roles will no longer exclude men from humanitarian assistance but will recognize them in the same way as women, children and elderly; the ones in need of aid.

The previous example of the UNCHR and UNICEF raises the question whether the gendered based norms in humanitarian organizations did change during the last years as Olivius (2016) argued, or that they are still rooted in categorizing men in the traditional gender beliefs and norms which exclude them from humanitarian assistance? The answer to this question is off great importance in International Relations (IR) because these gendered norms produce the sex-selective behavior of humanitarian organizations during their interventions. These interventions have to be based on the seven fundamental principles of humanitarian aid in which aid is only based on urgency and not on discrimination (OCHA, 2018). Humanity, impartiality, neutrality and independence are the core principles which humanitarian organizations claim to account for in every intervention. However, there seems to be a division in the literature. One side follows the argument of Carpenter (2003, p.661) that those principles are not being followed in reality. These scholars argue that humanitarian organizations have a gender bias which is responsible for excluding men in humanitarian assistance while earlier research already showed that of all war-affected non-combatants, adult men are being most at risk during a conflict. Despite this, typically women, children and

(5)

elderly have composed the ones who are helped first (Carpenter, 2003, p.662). The other side of the literature follows the argument of Olivius (2016). These scholars argue that humanitarian organizations no longer have a gender bias or gendered norms but instead also categorize men as the vulnerable ones who are in need of humanitarian assistance (Olivius, 2016, Dolan, 2014). Contrastingly, the example on UNCHR and UNICEF would hint at humanitarian organizations still not portraying and therefore recognizing men. Based on the contradicting literature and this example, this thesis will research to what extent humanitarian organizations use gendered norms in both their organization as well as in their assistance which might result in the exclusion of men. This thesis will therefore try to answer the following research question:

‘To what extent do humanitarian organizations use gendered norms which exclude men in their organizations and assistance?’

The answer to this research question will be provided by making use of a within case comparison study in which two typical non-governmental humanitarian organizations will be researched, both in general as well as in the case of Yemen. INTERSOS and Doctors Without Borders (DWB) will be used as typical cases. The organizations are categorized as typical because of their ‘typical' characteristics of a humanitarian organization such as being financially independent, having a global footprint and most important; providing humanitarian assistance. This thesis will first analyze to what extent the two cases in general display gendered norms. Additionally, this thesis will also research to what extent the organization's general behavior on gendered norms corresponds with their behavior in their humanitarian assistance. The case of Yemen will be used to illustrate this. Yemen is chosen here because this humanitarian crisis is being categorized as the worst crisis in years. If the humanitarian organizations behave in a certain way in such a crisis in which many civilians suffer and therefore the assistance has to be provided properly, one can extract this thesis's findings with great confidence onto other humanitarian organizations. Hence, Yemen is the first case, after the shift in the literature regarding also categorizing men as vulnerable, onto which humanitarian organizations could apply their ‘learned lessons’ from the past (Dolan, 2014; Olivius, 2016).

Researching INTERSOS and DWB will be done by making use of the analytical method of content analysis. Both imagery as well as text will be analyzed on the general homepage of the organizations, the ‘what we do’ and ‘who we are’ webpages and on the webpages which provide information about INTERSOS’s and DWB’s interventions in

(6)

Yemen. Hence, this thesis will use both primary and secondary literature from which the sub-codes will be derived which will be used for conducting the content analysis.

Scientific relevance

In the scientific field there is plenty literature written about the need for adopting women rights in humanitarian rights to account for the gender bias and gendered norms women face in humanitarian assistance, for example literature regarding gender based violence (Fox, 2001; Vann, 2002; Ticktin, 2011). However, scholars such as Carpenter (2003) gave the scientific field a wakeup call that not only women are the ones who face gender biases and gendered norms when it comes to humanitarian assistance. Many articles followed which all said something about establishing gender equality amongst men and women (Sjoberg, 2006; Wong, 2012; Helms, 2013). However, there is a gap in the literature regarding the application of these findings onto the ‘real world’. Literature does not seem to address whether the humanitarian organizations have changed or not, regarding the usage of gendered norms and biases which are responsible for excluding men in humanitarian organizations. Hence, many literature is written about UN-based organizations, just as Carpenter (2003) did, but not on non-governmental humanitarian organizations which this research will examine. In sum, this research will fill an important gap in the scientific field because its findings will add new insights on whether or not non-governmental humanitarian organizations also recognize men in their assistance. Or I will observe an empty shell in which humanitarian organizations claim to make need assessments based on urgency but in reality the organizations violate their core principles as a result of using gendered norms and biases.

Social relevance

The results of this thesis will also be relevant to the social field based on the following two arguments. First, if this research reveals that humanitarian organizations are not as gender-neutral as the organizations claim to be, based on the seven principles, their role as explicit norm-entrepreneurs fails. Second, the current situation in Yemen is devastating but at the same time it is a possibility for the organizations to provide assistance in the most proper way regarding gender beliefs and norms. The results of this thesis will therefore also add to the societal field whether or not Yemeni in need of aid are getting helped by the organizations or that they are being left behind based on gender.

(7)

Theoretical framework

There is currently a debate going on in which it is argued that politicization is taking place on humanitarian organizations (Barnett & Weiss, 2008). As mentioned in the introduction of this thesis, humanitarian organizations claim to be neutral in both their organizations as well as in their interventions. However, as Barnett and Weiss already argued back in 2008, the organizations and their aid are becoming more and more politicized which results in aid not being addressed to the people who are most in need of it. The scholars give as an explanation for this finding that humanitarian organizations are aware of the possibility that their interventions will get politicized, which could expose the dilemmas the humanitarian organizations face when assisting in a war-affected region (ibid.). As a consequence, humanitarian organizations are not willing to take the risk which causes the inability for the organizations to get the proper access. As a result, the humanitarian organizations have to compromise on neutrality, and due to this, the people who are most in need of aid are not addressed as Barnett and Weiss (2008) but also many others such as Dany (2015) and Barbé & Morillas (2019) stated. However, what is currently missing in this broader debate of the politicization of humanitarian organizations is the concept of gender.

To account for the concept of gender we have to address the phenomenon of gendered norms that are used in humanitarian organizations. These norms are in this case exclusively associated with women and children. On the one hand, you will find scholars who argue that humanitarian organizations are not as gender-neutral, particular in their interventions, as the organizations claim to be due to the exclusion of men in need for humanitarian assistance (Carpenter, 2003; Dolan, 2014; Rumbach & Knight, 2014). To examine whether or not this is still applicable to the cases of today and whether or not these findings also apply to NGO's, this thesis will use the theory of Carpenter (2003). On the other hand, you will find arguments which do not agree with scholars such as Carpenter. They state, primarily Olivius (2016b) but also others (Brown & Grävingholt, 2016; Kunz, 2014), that humanitarian organizations do account for gender and therefore can be characterized as inclusive and diverse. Olivius (2016b) describes three possible approaches of providing humanitarian aid and to what extent the approach accounts for gender equality.

I will first conceptualize the concept of gender and its meaning. After this, the theory of Carpenter will be presented. This will be divided into two parts, the first part focuses on addressing which gendered norms are at play on the organizational level of a humanitarian organization and which factors contribute to these norms being gendered and therefore also being biased against men. The second part is focused on the individual level and how the

(8)

gendered norms of the organizational level influence the individual level. Explanations for this relationship will also be provided. After the introduction of Carpenter's (2003) theory, I will first introduce the three approaches of Olivius (2016b) which humanitarian organizations can use. Then I will argue which approach suits best to use in this thesis. The chosen approach will then be used to examine if humanitarian organizations indeed do account for both men and women in their interventions as scholars such as Olivius (2016b) argues.

2.1 The concept of gender

This thesis will follow the conceptualization of gender by Carpenter (2003). One reason to follow Carpenter her conceptualization is because she refers to gender as something which is culturally constructed and which therefore regulates relations between and among men and women in various levels of social organizations (Carpenter, 2003, p.670). If this thesis would follow another conceptualization of gender, for instance the one which sees gender as biological characteristics, it becomes hard to examine whether or not humanitarian organizations behave in a gendered fashion towards ex- or including men because biological characteristics cannot explain this.

The conceptualization is based on the social constructivist approach which argues that norms are being defined as shared standards of appropriate behavior, held by a community of actors (Finnemore & Sikkink, 1998; Elgström, 2011). Hence, norms are ‘standards of behavior defined in terms of rights and obligations’ which is distinctive from rules (Krasner, 1983). The social constructivist approach is able to explain both the discourse and behavior of actors in the civilian protection networks (ibid.). Based on the social constructivist approach, humanitarian organizations believe that civilian immunity norms should be respected and, hence, they aim at a widespread implementation of those norms through purposeful action and persuasion. Those norms are constituted by gendered ideas which then constitute and give meaning to particular gendered concepts (Carpenter, 2003; Elgström, 2011; Finnemore & Sikkink, 1998). This means that gender beliefs will constitute gender norms that regulate appropriate relations between and among men and women; gender relations (Carpenter, 2003, p.671). One example of such an appropriate relation is the commonly used norm of men who are not allowed to hit women. This norm is being translated into the rule for men to don't hit women (ibid.). At the same time, gender beliefs could also exert a constitutive effect on the practices of a particular norm that then normalize and perpetuate a certain bias. One example is the ‘dress appropriately' norm, this norm is perceived as sex-neutral due to its applicability

(9)

to both sexes, but at the same time it works out for men and women differently and possesses therefore a gender bias.

These gendered norms and their biases are getting translated into the organizational identity of, for example, a humanitarian organization and will constitute gender roles (Hatch & Schultz, 2004). If this happens, certain norms are also getting transformed into the culture of an organization which will influence the decision-making and beliefs of both the organization as well as its employees (ibid.). As a consequence, it is hard to change certain norms, specifically when norms are already being gendered (Elgström, 2011). This therefore means that gendered norms affect both people's thinking as well as the ‘thinking of the organization'.

2.2 The exclusion of men in humanitarian organizations

Carpenter (2003, p.661) starts her article with a quote in which UNPROFOR General Morrillon communicates to UNHCR official Hollingworth, that men until the age of sixty will not be evacuated. This conversation between the general and official shows that based on age and gender, a particular group of civilians were getting excluded. This is contradicting to the guidelines which were set out in the UNHCR field manual (Carpenter, 2003, p.665). This stated that humanitarian assistance should be provided without making a distinction regarding sex, ethnicity or age. If situations occurred in which resources and opportunities were limited, the agencies should have used the reverse-triage principle. This principle accounts for impartiality due to the prioritization of the most vulnerable (ibid.). However, both the instructions given in the field manual and the agencies their main goal to protect civilians in a neutral manner were not met in the case of former Yugoslavia. Despite the knowledge that men and boys were the most likely to be executed, it were mostly women, children and elderly being evacuated during the humanitarian crises in Yugoslavia and not the most in danger; men (Carpenter, 2003, p.666). The only situation in which men between the age of sixteen and sixty did get evacuated was when they were used as part of a prisoner exchange (Carpenter, 2003, p.664)

Carpenter (2003) argues that the ‘regular' aid procedures and how the mass evacuations took place showed that the operational pattern of the humanitarian organizations in the Balkans did not go as impartial as the organizations claimed. Nothing seems to explain this phenomenon even when someone would follow the reasoning of the conventional wisdom in which women, children and elderly are perceived as the most vulnerable and men as combatants. If this would also work in practice, sex-selective assistance policies would be a

(10)

rational way of assessing whom to label as vulnerable. However, as Carpenter (2003, p.667) argues, there is no confirmation of these facts within especially war-affected regions. Women are not always the most vulnerable group of the population, left aside the sick, wounded and disabled ones. Despite this, this claim on women is often made and justified due to the lack of access to arms women have (ibid). This justification is wrong because there is no exclusivity to only adult men participating in hostilities, there are also older children and adult women who can fight alongside them. This means that neither the vulnerability of women nor combatant status can explain the exclusion of men by humanitarian organizations. What then does explain this phenomenon?

2.2.1 Sex-selective procedures and gendered norms

For Carpenter (2003, p.663) the proposed question could be answered when looking at the concepts of innocent civilian and especially vulnerable which are used by humanitarian organizations. Carpenter (ibid.) argues that the way in which gender ideas constitute those concepts will, in the end, have regulative effects on humanitarian organizations (ibid.). Due to this, Carpenter (2003) states that the taken-for-granted belief about men and women is socially constructed and has nothing to do with biological aspects of the sexes. This also applies to the gendered beliefs regarding women and children which are socially being categorized/constructed as noncombatants. Similarly, adult men are automatically perceived as combatants during war.

The civilian immunity norm (innocent civilian) is used by many humanitarian organizations. The organizations claim that this norm is a sex-neutral norm and protect those civilians who are not directly engaged in hostilities and therefore categorized as innocent (Carpenter, 2003, p.671). Even though the norm is categorized as sex-neutral, women and children are more often associated with the innocent civilian status rather than the combatant status in comparison to men which are more often associated with the combatant status rather than the status of innocent civilian as Carpenter (2003, p.670) in her analysis showed. This contradicts the argument that the civilian immunity norm is sex-neutral nut instead shows that there seems to be a bias against adult men. The cause of this bias against men lies merely in the modern laws of war which are traced back to the age of the Enlightment (ibid., p.672). During this time many writers of the laws of war agreed that the innocence of noncombatant men was a matter of argument but there was no doubt about women, children and elderly being treated as innocent. Based on this, women, children and elderly constitute the ‘other' group which are automatically being perceived as vulnerable. However, as Carpenter argues

(11)

(ibid.), these are assumptions made by idea-driven behavior and not reflected upon historical records of women's participation in combat. The only reason which gave women this status during a war is due to socially constructed gender ideas. The same argument holds for adult men which have been presumed as presumptive combatants and not as civilians, due to perceiving them as the ones most capable of bearing arms. Based on this gender belief men are being considered as dangerous and a legitimate target (Carpenter, 2003, p.672; Puechguirbal, 2012).

Another category which is of great importance in determining which group of the population needs aid is the concept of the especially vulnerable. Again, this concept is, just as the civilian immunity norm, argued to be sex-neutral. However, Carpenter (2003, p.674) states that in reality this concept is influenced through gender beliefs and therefore the norm is not as sex-neutral treated as the humanitarian organizations claim. This statement of Carpenter (2003, p.674) is based on a couple of findings. First, she highlights that there is a benefit for the humanitarian organizations when they are able to frame particular issues in terms of bodily harm to vulnerable populations. Meaning, if they frame particular groups as vulnerable they are better able to sell their programs to other actors. Due to this, the organizations will gain bargaining leverage over actors in the field with which they have to negotiate (ibid.). Besides this ‘personal' advantage for the humanitarian organizations, to frame as many groups as vulnerable, the concept of especially vulnerable is also influenced by making a decision solely based on age and sex rather than context. As a result, women, children and elderly are more often associated with the concept of vulnerability than adult men (ibid.). This assumption is not based on historical evidence nor on evidence which shows that men indeed are less vulnerable than women, children and elderly in war-affected regions.

So, the only evidence which could provide a just reason to the concept of vulnerability merely being focused on women is because women are made vulnerable during armed conflicts. Not due to their physical or psychological characteristics but due to the direct consequences of discrimination that women face (Carpenter, 2003, p.675; Puechguirbal, 2012). Additionally, Puechguirbal (2012) argues that women are labeled as victims by the international community irrespective of the increasing responsibilities women take over when men are absent. This also means that men are socially constructed, made, to fulfill the role of a combatant during war and therefore they are being excluded from perceiving them as vulnerable. They may physically be the least vulnerable group but due to the socially constructed male gender roles, men become far more vulnerable than women, children and the elderly. This gendered norm is also responsible for adult men being more exposed to greater

(12)

risks during war due to their socially constructed gender role as combatants (ibid.). One main contributor to this social construction of gender roles is language (ibid.). Language is one of the main factors which keeps the gender ideas and norms in place. For instance the previously mentioned framing of the concept of vulnerability.

2.2.2 Sex-selective procedures and policies

Besides the constitutive influence of gender ideas and their norms on humanitarian organizations discourse, also decision-making procedures in the field are needed to understand humanitarian organizations their sex-selective intervention policies.

Gender beliefs have a direct effect on the so-called ‘road maps' (cognitive maps) of protection workers (Carpenter, 2003, p.676). As Axelrod (2015) argued, people who have to make important decisions, are not necessarily aware of the shortcuts they use in order to make the decision ‘easier'. Merely in situations of incomplete information, people unintentionally make decisions based on, in this case, socially constructed gendered norms (Axelrod, 2015; Carpenter, 2003). Due to this, protection workers evacuate and help women, children and elderly significantly more in comparison to men because in their ‘shortcut' this group of the population is categorized as civilians (not combatants) and seen as vulnerable (ibid). These road maps are influenced by reliance on sex and age as proxy variables in the absence of clear criteria for making a distinction between civilians, combatants and not vulnerable and vulnerable. Hence, Carpenter (2003, p.677) states that this ‘logic of appropriateness', used by protection works in the absence of clear criteria, is not the only factor to blame in the sex-selective policies. There are two other reasons.

First, protection workers also rely on third parties and if those parties use particular concepts of gender, the protection workers will desire to maintain approval and legitimize their practices. This will only be successful if they use the categorization of concepts such as vulnerability in the same way as the third party does. In other words, the protection workers may have been aware in the Balkan of misusing the women and children first rule but due to the approval of others and socially inappropriate behavior they just followed the rule (ibid.). Second, even if actor B is aware of the gendered norms and its effects but actor A has bargaining leverage over actor B, actor B has to follow the gendered norms shaped in the policies of actor A (Carpenter, 2003, p.678). In the case of the Balkan, protection workers could have preferred to evacuate all civilians, including men, but if the belligerents or war-affected civilians themselves did not allow them to do so, it becomes impossible. Framing is therefore also an important aspect because actor A can frame a particular event in such a way

(13)

that B has to adjust to this frame. Autesserre (2009) described this phenomenon of framing as the opportunity in which actor A can authorize and justify specific practices and policies while precluding others such as the gendered norms. If there is one actor in the field, for instance in the Balkan the BSA, who categorizes men as combatants rather than civilians, this will pose an intractable barrier to not only providing aid to men but also to evacuate them (Carpenter, 2003, p.678). As a consequence, decision-making procedures also contribute to the exclusion of men.

Concluding, this chapter started with posing the question of how one of the core principles, impartiality, was not met during the humanitarian crises in the Balkan. The answer provided by Carpenter (2003) is that gendered norms does not only influence the moral framework in society but also influence how humanitarian assistance is carried out. Hence, Carpenter (2003) showed that the norm of innocent civilian and the concept of vulnerability are responsible for the exclusion of men in humanitarian assistance. Both done in- and intentionally by humanitarian organizations as well as by actors such as the BSA who merely have traditional gender ideas and therefore will always see men as combatants. Hence, Carpenter (2003) in the end argues that it is almost an obligation for the humanitarian organizations to correct this gender bias.

Based on this theory I will argue that even though Carpenter provided this knowledge and recommended to correct the gender bias in humanitarian organizations, it did not happen. One reason to state this is because Carpenter showed that gender beliefs are strongly embedded into the society and hard to change even though it seems that the general belief of the society became more diverse over the last years. Second, Carpenter addresses that the exclusion of men is also a result of the gender beliefs which the war-affected, and other actors in the particular war region, have. Most countries who have to deal with humanitarian crises still hold on to traditional beliefs, ideas and values and this also applies to their gender beliefs. Hence, Carpenter showed that humanitarian organizations need to sell their programs to gain bargaining leverage over other actors. Due to this and the knowledge that those ‘traditional countries' will view men as combatants, humanitarian organizations will not incorporate men into their interventions because it will not stand a chance. Merely based on the argument that gendered norms are hard to change, I argue that humanitarian organizations still use the traditional gender roles which cause the exclusion of men from humanitarian assistance in this day and age. Based on this the following proposition is constituted for this thesis:

(14)

2.3 Critiques on the exclusion of men

Olivius (2016a, p.2) argues that humanitarian policy and practice did change the past years. Men are being recognized by humanitarian organizations which becomes visible in the change of policy terminology which used to refer to ‘women’ but has changed into referring to more general terms such as ‘patients’ or the ‘ones who are most in need of help’ when talked about whom to provide aid to. Olivius (ibid.) poses the question if this is a welcome shift, as Carpenter suggested, towards recognition that gender has multiple forms of vulnerability or that this change shifts the attention and resources away from the yet unfinished struggle for women’s rights.

In her article she theorizes that, contrastingly to Carpenter, it is problematic to get adult men represented in humanitarian policy and practice because this will limit the usefulness of gender equality and liberation (Olivius, 2016a, p.8). This limitation on gender equality will occur because the inclusion of men will obscure power relations that are constitutive of gender differences. Put it differently, it is not desirable to include adult men in humanitarian policy and practice. This will only lead to a situation where the needs of men and women are taken into account equally rather than transforming the unequal relations of power between the two (ibid.). It will eventually lead, as Olivius argues, to a situation in which women will be excluded because they will no longer be categorized as vulnerable. If this happens, you will end up with the same situation as a couple of years ago before the feminists fought for equal rights. Therefore Olivius (2016a) argues that the target of correcting gender biases should not be men but instead women because the old struggle is not yet finished. After solving this problem, the social order should become the target of change to solve the overall problem of the exclusion of both men and women (Olivius, 2016a, p.15).

However, as Carpenter (2003) showed, merely norms and gender beliefs are socially constructed in a way that causes biases for men regarding the concepts of civilians and vulnerability in humanitarian assistance. Hence, Carpenter mentions that her ‘promotion' of getting men included in humanitarian policy and practice, does not neglect the idea that women, children and elderly are vulnerable during war. It seems like Olivius is merely criticizing this point because she speaks of masculinity studies, such as Carpenter, causing a shift away from the unfinished struggle of women. Hence, Olivius (2016a, p.16) also argues

(15)

that getting men categorized as vulnerable will be destructive for the existing masculine roles. However, as Carpenter (2003) argues, in times of war every civilian should get equally treated in the assessment of being evacuated or helped or not. In times of war or big humanitarian crises, one should not care, as I argue, for masculine roles getting less masculine. Hence, Olivius only mentions in her critique the problem with the overall structure and does not mention the specific problem; the socially constructed gender beliefs and norms which influence the behavior of humanitarian organizations.

2.4 The recognition of men in humanitarian organizations

After Carpenter published her article, many things have changed in international policymaking. The humanitarian organizations and their aid are no exception to this (Olivius, 2016b, p.1), for example the gender Handbook for Humanitarian Action published in 2017 by IASC. The goal of gender equality is widely adopted in humanitarian policy texts, field handbooks and even the gender mainstreaming strategy has been adopted by the United Nations (UN) and non-governmental organizations (NGOs) as Olivius (ibid.) argues. The main reason for those agencies to do so is because they are aware of fulfilling the humanitarian imperative to save lives, and relieve suffering in times of war and disaster for every civilian in need of aid (ibid.). However, Olivius (2016b, p.1) at the same time mentions that it is still not very clear what precisely the concepts of ‘mainstreaming gender' etc. mean in humanitarian field practice. Therefore she analyses humanitarian gender handbooks and how they interpret gender in their humanitarian policy and practice. Hence, she provides three different approaches and how each approach accounts for gender equality based on possible limitations or advantages. I will introduce all of the approaches and then pick the approach which is the most relevant for this thesis.

2.4.1 The basic needs approach

The first approach which humanitarian organizations use according to Olivius (2016b) is the basic needs approach. This approach is influenced by the classic humanitarian imperative to save lives and reduce suffering. The main goal is therefore to ensure that all civilians in need of aid receive the necessities of life, derived from the handbook by IASC (2006). Hence, this approach does take women, children, elderly and adult men in account but only ensure that the basic needs are met (Olivius, 2016b, p.9). In practical programming, this approach results in merely focusing on getting equal numbers of men and women in a particular aid program (Olivius, 2016b, p.6; Brown & Grävingholt, 2016, p.192). One interviewee of an NGO in

(16)

Thailand said that they have to use quotas that have to be filled with equal numbers of males and females in order to ensure gender equal access to services, resources and evacuations possibilities (ibid.). Brown and Grävingholt (2016) would add to this that equal access also consists of men and women equally being able to benefit from developments that are carried out by humanitarian organizations. Hence, Olivius (ibid.) also mentions just like Carpenter (2003) did, that humanitarian organizations need the support of humanitarian donors and these donors reinforce the focus on equal access in terms of numbers. However, these donors also cause ‘chaos' as Dunn (2012) argues in which it becomes complicated for humanitarian organizations to provide aid to the most in need due to restrictions caused by these donors which is a limitation of this approach.

The concepts of civilian protection and vulnerability are central to the basic needs approach (Olivius, 2016b, p.7). Based on vulnerability, the humanitarian organizations decide which civilians require humanitarian assistance (ibid.). However, Olivius (2016b, p.7) argues that due to the use of the concept of vulnerability and not the concept of gender, much emphasis is placed on women’s vulnerability. Therefore Olivius (2016a) argued in her critique on Carpenter (2003) that the focus on getting men included will in the end not lead to gender equality. This is the result of humanitarian workers in practice having to make need assessments based on the concept of vulnerability which will result in emphasizing women’s vulnerability more. She argues that this has nothing to do with biased gender beliefs but just with the circumstances in which women find themselves, think about sexual and gender based violence (SGBV) (Olivius, 2016b, p.7). Despite this, Olivus (2016b) argues that this approach has one major advantage; the approach is aimed at immediately identifying discrepancies between men and women and accounts for this. This tracking of discrepancies and yet ‘fill' them with equal numbers of men and women is the first step towards gender equality but not towards accounting for gendered norms in humanitarian organizations and their assistance. In practice, protection workers will always have to make a decision based on gender by counting the numbers of men and women in a particular aid program. The understanding of gender inequality remains therefore unlimited and not addressed in humanitarian organizations.

2.4.2 The instrumentalist approach

The second approach which accounts for gender and corrects the bias in humanitarian aid is the instrumentalist approach. This approach rests on an understanding of gender as differences between men and women in which the different sexes respond differently to emergency and displacement (Olivius, 2016b, p.8). As a consequence, aid will be the most

(17)

effective in humanitarian programs when it understands those differences and takes them into account. Hence, this means that this approach takes the qualities and capacities of both men and women into account to be able to provide aid as effective as possible (ibid.). One example for this are the humanitarian policy texts which describe women in times of war as the secret weapon to beat hunger. Another example for this are the policies that promote localized women-led approaches in times when men are absent (Yermo, 2019).

The main advantage of organizations that provide aid according to this approach is the largely positive and active images of both men and women that it conveys as Olivius (2016b, p.10) states. For instance, the active image of women as weapons to beat hunger will, in the end, reduce the focus on women as victims of war and as the most vulnerable group (ibid.). As a consequence, humanitarian organizations will therefore also be able to perceive adult men as vulnerable in situations in which this is the case. However, this approach does not transform traditional gender relations and power dynamics in pursuit of correcting gendered norms which cause gender inequality. It only focuses on the differences between gender (ibid). By doing this the possibility arises that the counter effect of this approach is that the gendered norms, which constitute gender roles, are kept. Due to this possibility, this approach will not naturalize and reaffirm gendered norms because it keeps recognizing men and women in their traditional gender roles. However, one could also argue that this is the exact reason why this approach would be effective because it recognizes the sexes in their traditional gender roles which are almost always used in the war-affected countries in which humanitarian crises are going on. If humanitarian organizations therefore use those traditional gender roles, their bargaining leverage will increase because the society in which the organization intervenes ‘respects' the gender roles. Due to this all civilians, including adult men, could be addressed.

2.3.3 The modernization approach

The modernization approach is the last approach (Olivius, 2016b, p.10). This approach is based on two understandings. I would argue that Olivius follows in her first understanding the modernization theory by Inglehart and Norris (2003). It looks like she labeled this last approach purposely in the same way as the theory. Olivius (2016b, p.11) argues in the same way that, in her case humanitarian organizations, went through the process of modernization which occurred because the society in which the organization is based went through the same development. As a result of this process, the society, and its institutions such as the humanitarian organizations, have adopted and institutionalized modern liberal democratic

(18)

values. Hence, these values make, merely western, societies more modern, liberal and democratic in comparison to the crisis-societies in which the organizations intervene because the latter did not go through this process of modernization (Olivius, 2016b, p.10; Inglehart & Norris, 2003, p.10). This modernization brings systematic changes in gender roles, as Inglehart and Norris already argued (2003, p.10), towards more gender egalitarianism. This means that in the last years the humanitarian organizations went through a process of change, alongside the cultural and industrial change as Inglehart and Norris (2003) would argue, from gender-role and belief traditionalism towards more modern ones. However, Olivius (2016b) does not explicitly mention this link with the modernization theory. Yet, I think it is important to connect her approach with this theory because then her argument of humanitarian organizations being nowadays based on modern liberal norms and values, and therefore modern gender beliefs, is more acceptable.

As argued, Olivius’s last approach is based on her claim that nowadays humanitarian organizations are based upon modern, democratic and liberal values (Olivius, 2016b, p.11). Modern liberal values go hand in hand with modern liberal gender beliefs and norms as Oláh and others (2014; Jokinen & Kuronen, 2011; Knight & Brinton, 2017) showed. Exactly these norms have changed over the last years into both men and women ‘getting’ new gender roles which are more diverse and egalitarian and no longer stochastically attached to one sex (ibid.). Due to this change over the last years in the liberal norms and values, family trends and patterns have been paralleled by changes in gender roles (Oláh et al, 2014, p.2). Female roles have been expanded towards also including, for example, economic provision for a family. Hence, male roles have become more inclusive towards the involvement in family responsibilities (ibid.). Men are being more and more recognized in the role of a caretaker and women as breadwinners. Hence, the Male Role Norms Inventory (MRNI) of 2014, once invented by Levant, which for instance measures if men still categorize themselves as the ones who have to fight when things get tough, also showed that traditional masculine ideologies are becoming more egalitarian when comparing them with the MRNI of 2007 (Levant, R. et al, 2014, p.9). The norm of only men being the ones capable of fighting is changing towards a gender belief that also women are capable of doing so (ibid). This means that modern gender norms are becoming more relaxed and fluidly interchangeably of gendered responsibilities.

Here the second understanding of Olivius (2016b) continues which consists of two parts. The first part explains why and how the humanitarian organizations have adopted these modern gender roles in both their organizations as well as in their assistance (Olivius, 2016b,

(19)

p.11). As a result of Olivius’s claim that nowadays most humanitarian organizations have adopted modern liberal values, the organizations are capable of recognizing the gender biases which come with the traditional gender roles. The organizations will account for those and transform them into the modern gender roles which are recognized by liberal values (ibid.). For example, the organizations are aware of the bias that men are more often seen as combatants and therefore less as a victim. Instead of working with this bias, the organizations recognize this bias and account for it (ibid). Due to this, humanitarian organizations will recognize every civilian in their interventions as a possible victim which is in need of aid.

The second part is a result of the first part of Olivius’s (2016b) argument. Because the humanitarian organizations have incorporated the modern liberal gender beliefs and norms in their organizational identity, the organizations automatically carry the responsibility to implement this in the war-affected societies. This responsibility is derived from what Olivius (2016b, p.11) calls ‘the window of opportunity’ and her categorization that war-affected communities, or societies, are less developed, traditional and backward. A humanitarian crisis will cause a window of opportunity for the humanitarian organizations to promote and develop gender equality in the crisis-affected society because the society in which the organization intervenes still uses the traditional gender roles which have to be transformed into the modern gender roles. Transforming gender beliefs can only be done by addressing the social and cultural structure which can only be changed in situations of emergency and displacement. Hence, Olivius (2016b, p.13) argues that this will be translated into programs that are focused on implementing the liberal gender roles, beliefs and values in order to also implement gender equality and recognition in the crisis-affected society. One way of indicating if a humanitarian organization uses modern liberal norms and values and therefore does include and recognize men, is looking at which kind of programs they implement in the country they intervene, as Olivius (ibid.) claims. The time-frame of this last approach is therefore also different due to its focus being not only at the immediate crises situation but instead, it also focuses on the long term. Hence, the primary goal is more far-reaching and different in comparison with the earlier mentioned approaches, having a time-frame which is focused on not only the immediate crises.

If humanitarian organizations therefore rely on this modernization approach, they will seize the opportunity to lay a foundation for teaching new norms and values which are often absent in those war-affected countries. Therefore this approach connects the promotion and recognition of gender equality with a broader project of liberal peacebuilding and state building (Olivius, 2016b, p.11). In practice, this means that humanitarian organizations often

(20)

inform programs geared towards changing those civilians their gender beliefs. This is different from the previous discussed approaches which tend to see gender differences as something which is fixed and attached to individuals. The modernization approach links gender with the structural understanding of power. Olivius (2016b, p.12) states that by recognizing this in humanitarian organizations, the organizations will account for all civilians because the underlying power dynamics are understood. She argues that if humanitarian organizations do not understand this, diversity and therefore the inclusion of both men and women in the humanitarian organizations will not happen. Olivius (2016b) argues that nowadays more humanitarian organizations are working as the modernization approach prescribes and therefore they do account for both men and women.

As a consequence, I argue that the third approach is the most relevant for this thesis because it can examine whether or not humanitarian organizations account for gender-neutrality and not the ‘first phase' of gender-equality which is promoted with the first approach. The second approach merely focuses on the instrumental biases carried out, or not, by humanitarian organizations. To change these biases in an organization, you should first research whether the structure and culture of the humanitarian organization use gender biases at all. This can be researched with the third approach. Based on the modernization approach, the second proposition of this thesis is:

 Men are no longer being excluded in humanitarian organizations. 2.5 Reflection

Even though this thesis critically looked at which theories would suit best in order to answer the research question, the theories are not completely perfect. Some side notes have to be made and some possible pitfalls discussed before we move onto the methodological chapter. First of all, Carpenter's theory is used in this thesis to explain the exclusion of men in humanitarian organizations, in specific humanitarian organizations which are categorized as non-governmental organizations. However, Carpenter did her research on UN based humanitarian organizations, such as the UNCHR. This means that the unit of observation in both pieces of research are different and therefore it could be that some of Carpenter's findings will not be found due to the organization being different. Second, the two propositions of this thesis are posed in a way that either there is an inclusion of men nowadays in humanitarian organizations or there is not. There is no expectation that there would be a hybrid form in which men are a bit included but not fully. However, there is a chance that this in reality could be the case and as a consequence one of the other two approaches, which will

(21)

not be tested, are met. Hence, the theory of Olivius is merely based on getting both sides of the sexes equally included in humanitarian organizations and therefore her theory is more about how humanitarian organizations should act. Due to this, her theory could be characterized as more normative, in comparison with Carpenter's theory.

Besides these points of attention, the chosen theories also have some strong points which will differentiate this thesis from other research. First, both propositions, derived from mainly Carpenter and Olivius, claim another outcome regarding the inclusion of men in humanitarian organizations. This means that both sides of the discussion, are men indeed getting included or not in humanitarian organizations, can be researched. Second, both theories are based on actual interviews and therefore one could argue that these theories are based on ‘reality'. What is really going on in these humanitarian organizations, and not based on something which we ‘think' is going on in these organizations. Lastly, by using the social constructivist approach, as Carpenter did, to explain how gendered norms are getting included in not only society but also organizations, this thesis will provide more than only findings on the inclusion of men. It will also be able to provide an expectation on why men now indeed are getting included, in contrast with Carpenter's findings, or why they are still not getting included. This will then provide a starting point for further research.

In sum, both Carpenter’s (2003) core findings as well as Olivius’s (2016b) findings are used to illustrate the two sides in the literature. On the one hand, the scholars who argue that men are still being excluded in humanitarian organizations and assistance as a result of gendered norms which cause traditional gender roles. On the other hand, the scholars who argue that men are nowadays getting included and recognized by humanitarian organizations as a result of the modern gender roles which no longer use the gendered norm of merely combatant which causes the exclusion. However, due to the scope of this thesis, I will not research Carpenter's arguments on the third parties. Researching this could be a research on its own. Hence, I will not intensively look at the cognitive roadmaps and shortcuts which Carpenter theorized because this has to be researched by conducting interviews with humanitarian workers and this is not the main focus of this thesis. However, if it is possible to add some additional findings on this I will add them if it supports or contradicts some of my main findings in the empirical chapter. I will also not research the basic-needs and instrumentalist approach of Olivius (2016b) as I argued before. In the next chapter, I will provide information on the method I will use and which cases I will look at.

(22)

Methods

In this chapter, I will argue why I will make use of a qualitative way of doing research, a case study method and why I have chosen to use content analysis to answer the research question. This chapter will also provide information on the selected typical cases of INTERSOS and DWB. Hence, I will make an operationalization that will get translated into the coding tree of this thesis.

There are several techniques for doing research (Kothari, 2004). To decide which technique this thesis will use, the first decision has to be made regarding whether a quantitative or qualitative way of doing research is used. Quantitative research is aimed at measuring a particular quantity or amount and is merely applied to phenomena that can be expressed in terms of quantity (Kothari, 2004, p.3). In contrast, qualitative research is merely based on discovering underlying mechanisms that cause, or explain, a particular phenomenon (ibid.) Based on this, this thesis will do a qualitative research to discover to what extent humanitarian organizations use gendered norms in their organizations and interventions. Quantitative research will not be able to research this because researching norms cannot be expressed in quantities.

Yet, this thesis will use the case study method which is a popular form of qualitative research (Kothari, 2004, p.113; Gerring, 2004, p.341). Using the case study method involves observations of a social unit that is focused on studying the particular case more in depth rather than breadth (Kothari, 2004, p.113). Hence, this method focuses on researching processes and interrelationships of the particular phenomenon (ibid). This thesis will therefore provide a case study in which I will intensively look at two non-governmental humanitarian organizations within the case of Yemen. One benefit of doing a case study is the ability to apply the above discussed theories onto situations in the ‘real world'. Hence, while doing a case study, rather than for instance a quantitative study, the general phenomenon, regarding gendered norms in humanitarian organizations, can be understood by looking first at a smaller

(23)

case (Gerring, 2007, p.79). One major drawback of doing a case study is that it is merely based on the interpretation of the researcher. However, this thesis is focused at researching a ‘real world’ phenomenon and therefore it needs a human’s interpretation to get to know what humanitarian organizations really say and not what the organizations want you to know, which would be the case with a quantitative study based on data sources merely provided by the organizations themselves. A quantitative study would for instance only count the frequency of men getting displayed, or not, on the websites of the humanitarian organizations, what this then entails would remain unclear.

3.1 Case selection: INTERSOS and Doctors Without Borders

The findings of this thesis will be used as a starting point for further research. To really provide findings that will add ‘new' knowledge on the consequences of using gendered norms in humanitarian organizations, and in order to generalize these findings onto other cases, I have to research a typical case. Preferably, researching two typical cases and compare them within a case to research whether their general behavior holds, or changes when their gender beliefs have to be applied to ‘the real world'. I will use two typical cases due to the following argument; if the core theories of this thesis are applicable onto the organizations and the within case of Yemen, we can expect that these findings can be extracted onto other cases in which humanitarian crises are at stake and humanitarian organizations provide aid. If this thesis finds that the humanitarian organizations do not account for men, based on the first proposition, we can assume that other NGO's will behave in the same way in the same case but also in other cases which are categorized as typical. This argument also holds for the second proposition which posed that men are being recognized and accounted for by humanitarian organizations. If the findings of this thesis prove that humanitarian organizations do account for men in their organizations and interventions, we can expect that other organizations will act in the same way. As a consequence, I will research two typical non-governmental humanitarian organizations and do a within case comparison study. In order to do a within case comparison with typical cases, the two humanitarian organizations have to be both typical and have to intervene in the same case/country. However, as Snyder (2018) showed, there are many non-governmental organizations around the world and much humanitarian crises in which they intervene. The difficulty is which ones to choose and within which case, therefore I will turn to Gerring (2007).

Gerring (2007, p.89) has constituted multiple techniques of selecting between the different types of cases which all are used differently in doing research. Due to this thesis

(24)

researching whether or not humanitarian organizations use gendered norms which are responsible for the in- or exclusion of men in both their organizations as well as their interventions, the cases have to be typical in order to generalize the findings. A typical case is a case that is representative, as Gerring (2007, p.89) argues, for a broader set of cases. Some of those cases serve an explanatory role in which the researcher chooses a case based on a set of descriptive characteristics and then searches for causal relationships within that case. Other cases serve as a causal model of a particular phenomenon, due to this the researcher can investigate the identified outcome (Y) (Gerring, 2007, p.89). To research this outcome, the researcher chooses a typical case that represents the causal relationship. In both roles, the typical case has to have some typical descriptive characteristics which represent the identified outcome or causal relationship. This thesis will use the typical case in order to look for the causal relationships as theorized by Carpenter (2003) and Olivius (2016b), respectively the relationship between the presence of gendered norms in humanitarian organizations and to what extent this results in the in- or exclusion of men. To do so, the humanitarian organizations have to be typical and the within case has to have ‘typical circumstances' of a humanitarian crisis as well.

The typical cases will be respectively the humanitarian NGO's, DWB and INTERSOS. They both have core characteristics of humanitarian NGO's: a global footprint which means that the humanitarian NGO has a global presence in both developing as well as industrial countries (Jayawickrama & McCullagh, 2009). Hence, a long-term presence which means that the humanitarian NGO is present in the country in which they intervene over a long period and not only for the time in which the aid is acute (ibid). Independence is the third characteristic which refers to the humanitarian NGO being independent of governments and other institutions. This characteristic goes together with the characteristic of the diversity of funds instead of funds which are released by governments. The last characteristic is the values base which means that humanitarian NGO's are guided by a mission, vision and principles and not only by project goals (ibid.). All of these typical characteristics are present in the humanitarian organizations Doctors Without Borders and INTERSOS. Yet, both organizations are currently operating in the biggest humanitarian crisis of this time: Yemen.

DWB describe themselves as an independent, neutral and impartial global movement providing medical aid where it's needed most (DWB, 2019). The organization consists of both doctors and other health sector workers who can provide lifesaving medical care to those most in need of it. The organization is merely funded by donations and uses 89.4% of these funds for their programs. One of these programs is aimed at helping the civilians of Yemen who are

(25)

living in a humanitarian crisis and therefore are not able to properly access healthcare. DWB started this operation back in 2015 and already treated more than 973.00 patients (DWB, 2019). The other organization is INTERSOS, again a non-governmental organization that is not controlled by any political, national or international authority. INTERSOS their main goal is to provide aid while at the same time involving the local population when implementing or carrying out particular actions (INTERSOS, 2019). In 2008 they began their operation in Yemen. It started with merely assisting refugees in the Kharaz and Basateen camps and identifying victims of human trafficking. When the conflict in 2015 began to worsen INTERSOS continued to provide humanitarian aid. Since 2015, INTERSOS’s assistance is merely focused on helping displaced persons and refugees who want to flee ongoing clashes and bombings (ibid.). Hence, INTERSOS provides aid in the most remote and hard-to-reach locations in Yemen.

Both organizations are purposely chosen to study due to a couple of reasons. First, both organizations are characterized as NGO's which means that Carpenter's core findings will get tested onto two cases which do not have the same characterizations as the UN humanitarian organizations Carpenter (2003) researched. Due to this, this research will add new insights to the already existing literature regarding the extent to which humanitarian organizations use gendered norms and as a consequence in- or exclude men in their organizations and assistance. One characteristic which is different between humanitarian NGO's and UN-based humanitarian organizations is that NGO's have a non-political foundation. This might result in a more diverse organization with more gender diverse gender beliefs. Hence, the programs of INTERSOS and DWB rely on funding of private and public donors, due to this their way of working has to match with the general (gender) beliefs of their donors, otherwise they will no longer donate. Yet, due to INTERSOS and DWB being more independent they can operate easier in particular areas compared to UN organizations which always have to ‘ask' permission. Researching non-governmental humanitarian organizations could show us whether or not Carpenter's theory is also applicable on NGO's. The findings of Olivius (2016b) were already based on NGO's. The second reason to use DWB and INTERSOS as units of analysis is that both organizations operate in Yemen but with different goals. On the one hand, DWB is merely focused on providing medical aid and INTERSOS is merely focused on helping civilians to flee dangerous areas. This means that the propositions will be tested on different mandates. Researching INTERSOS and DWB also makes it possible to look at whether there is a difference in using gender biases regarding different aspects of humanitarian interventions; healthcare, distributing essential goods, protection and

(26)

evacuation. This would not be the case if I had chosen to research two NGO's which, for example, both provide medical care.

Yemen is chosen in this research as a typical within case due to two reasons. First, Yemen has some core characteristics of a ‘typical humanitarian crises'; hunger, a higher number of deaths, a displaced population, human rights violation, barriers to care and poor security. All of this is caused by both a man-made crisis, between the Saudi-led coalition and the Houthis and natural disasters such as the cholera outbreak (DWB, 2019). The second reason why Yemen is chosen as a typical case is that it is the first humanitarian crisis that has all typical ‘ingredients' of a humanitarian crisis; it is a man-made crisis as well as a natural disaster which the cholera outbreak shows. It is therefore relevant to use Yemen as a case because the findings of this thesis will also apply to ‘less extreme' typical cases.

3.2 Method: Content Analysis

Analyzing a typical case can be done with several techniques. For example, using the method of process tracing, or doing a content or discourse analysis. To really grasp whether or not DWB and INTERSOS in- or exclude men in their organizations both pictures and words have to be analyzed. This can be done by using the analytical method of content analysis, which will be used in this research.

There are three approaches to content analysis. The first one is the conventional approach which is generally used in a study design that is aimed at describing a particular phenomenon (Kondracki & Wellman, 2001). This is merely done when there is literature written about a certain phenomenon but the literature does not address a particular variable or concept which seems to be at play in the phenomenon (ibid.). This approach is also used to identify new insights from already existing literature. Therefore this approach is merely focused on inductively developing categories that could provide some new insights. Inductively developing codes means that documents, data, are read word by word to in the end derive codes from it because these words, as the social constructivist approach also argued, capture key thoughts or concepts (Kondracki & Wellman, 2001). Hence, the codes are divided into different subcategories and there is an ability with this approach to combine different codes which are related or linked with each other.

The second approach is the directed content analysis in which deductively, codes are constructed (Potter & Levine-Donnerstein, 1999). This approach is used when the research goal is aimed to validate or extend an already existing theoretical framework or theory. The difference between this approach and the former is the more structured process this way of content analysis follows. This means that codes are derived from already existing theories, or

(27)

prior research, by identifying some key concepts from the theory as initial coding categories (Potter & Levine-Donnerstein, 1999). With this approach, codes are established by first intensively getting into the theory (or theories) and highlighting all text, or words, that appear to represent the phenomenon that is being researched. After this, the highlighted passages have to be translated into codes. Codes that are not belonging to a particular category, based on the theory, can be given a new code immediately or can be identified later to determine if they represent a subcategory of the other codes, or if it represents some new information.

The last approach to content analysis is the summative content analysis (Potter & Levine-Donnerstein, 1999). This way of doing content analysis starts with identifying particular words or content in the text. The goal of this approach is to understand why certain words are used in a particular context. Put it differently, understanding the usage of particular words in order to grasp the exact meaning of the content. This is also referred to as the manifest content analysis, which is merely based on counting the frequency of certain words (ibid). One could therefore argue that this approach is more focused on doing quantitative research in comparison with the other two approaches. However, the summative analysis is not merely quantitative because it also relies on including latent content analysis. This refers to the process of interpretation of content, in which the researcher is going to look at the underlying meanings of the used words or the overall content (ibid.). The summative approach therefore has two parts, the first part consists of counting particular codes (more quantitative) and the second part consists of interpreting these codes. One major drawback of this approach is that there is no reliance on theory. The researcher is only able to find the relation between the word frequency and his/her own interpretation of a particular code.

This research will merely use deductive coding and therefore the direct approach of doing content analysis. If needed, the codes, derived from the deductive content analysis, will get supplemented by doing inductive coding as well which means that I will combine both the approach of deductive, direct, content analysis as well as the inductive, conventional approach, of doing content analysis. Hence, this thesis will make no use of the third approach of doing content analysis. To answer the research question properly, the posed propositions have to be dis- or confirmed and these propositions are derived from theories. This means that the only way to do this is by constructing codes which are derived from these theories. Therefore it would make no sense if only the relation between word frequency and interpretations of these words gets researched. Other research methods that are frequently used in a case study, such as interviews would not apply in this thesis due to the following reasons. First, time limitations, costs and resources are three factors that make it not possible

The scale of exclusion