26-31 May 2002
Gender Integration in the Armed Forces:
A Cross National Comparison of Policies and Practices in NATO
ISCTE/University of Lisbon
European University Institute - Florence
Working paper: please do not quote
for Professor Gerard J. DeGroot
Dr. Helena Carreiras is assistant professor of Sociology
at the University of
Lisbon (Instituto Superior de Ciencias do Trabalho e da
Empresa) since1989. She is also a researcher at the Centre
for Studies and Research in Sociology (CIES). During the
past decade she has developed extensive research in the
fields of Political Sociology, Armed Forces and Society
and Gender Studies. In 1994 she concluded a master's thesis
on the participation of women in the Portuguese Armed
she is concluding her PhD thesis, 'A comparative study
of the participation of women in the Armed Forces of NATO
countries', at the European University Institute in Florence.
We of course have a particular interest in this last one
and are all anxious to ear your on results. I also understand
that this was not an easy task to do in getting the different
nations to participate in the submission of data. Nevertheless,
Dr. Carreiras persevered and was able to complete her
Among other publications she is the author of 'Women in
the Portuguese Armed
Forces' (1987) and 'Women in arms: The military participation
of women in Southern Europe' (2002)'. Without further
do, Dr. Carreiras, the floor is yours.
Dr. Carreiras has agreed present her research results
on 'A comparative study of the participation of women
in the Armed Forces of NATO.
A common observation regarding the current participation of
women in the armed forces in western democratic societies is
that there is great variation regarding the extent to which
the different NATO countries have integrated women. Such variation,
ranging from an almost total absence, rank limitations, segregated
training and severe functional restrictions, to relatively open
career patterns, full integrated training and access to combat
roles, is due to complex constellations of factors that in each
case have differently affected policies and practices.
However, despite this strong heterogeneity regarding integration
policies, all 18 NATO countries (2)
have admitted and generally increased the number of women in
their Armed Forces. During the past three decades various restrictions
have been eliminated; women have progressively been allowed
to military academies and obtained access to a wider variety
of positions and functions, including those related to combat.
Drawing on existing theories regarding the multiple factors
that affect women's military roles, in this paper I start by
looking at various indicators regarding the position of military
women in the various countries. Comparisons will rely, among
other dimensions, on global representation and occupational
and hierarchical distributions. The goal of this analysis is
to create an 'integration index' regarding the participation
of women in the military. This index will then be contrasted
against some of the selected independent factors that have been
proposed as determinants for the role women play in the armed
forces, in order to evaluate the impact of the various factors
that may influence gender integration in the military.
Factors affecting women's participation in the military: an
Women's military participation is affected by a wide variety
of factors - deriving from military organisational features,
social/economic structure and cultural and political values.
In one of the very few essays aimed at explicit theory building
in this field, Mady Segal has outlined a systematic theory of
the conditions under which women's military roles may expand
or contract, assessing the relative and combined effects of
those variables (Segal 1995).
Drawing on the author's proposal as well as on other theoretical
and empirical inputs, I have built an analytical model regarding
the various factors and conditions affecting the integration
of women in the military. Due to existing limitations to the
size of this paper I will not develop the various aspects here
(3), but rather refer to the model
and its different components while testing the specific hypothesis
on the impact of social, political, economic and military factors
Figure 1 - Analytical Model
The position of women in military organisational structures:
representation and integration policies
In the year 2000 around 280 000 women served as volunteer soldiers
in 17 NATO countries (4). Table
1 shows the absolute numbers and percentages of women in each
country, as well as its evolution.
(All following figures are included into .PDF/206Kb
From the point of view of service distribution, women's representation
is higher in the more specialised services: Navy and Air Force.
Women make up for 6.3% of Army personnel, 10,2% of the Navy
and 12.2% of the Air Force.
Occupational distribution and rank structure
The occupational distribution of military women reveals a strong
cross-national segregation pattern. Data available for 13 NATO
nations (5) shows that more than
two thirds (70%) of military women are concentrated in support
(personnel/ administration/logistics) and medical functions,
17,5% in technical areas (engineering, communications) and only
7% occupy positions in the more operational areas in combat
arms (artillery, infantry, cavalry).
Even if the relative weight of each of these different types
of functions in the overall military organisational structure
is taken into account, women are clearly over represented in
the traditional female areas. In fact, support functions in
the areas of personnel, finance, administration and medical
services employ an average of 46.1% of military personnel in
NATO countries, against more than 70% in the case of women.
The opposite happens in technical and particularly operational
functions where the asymmetry is also extremely visible: while
22.2% of military personnel occupy operational functions only
7% of women do (Figure 4).
As far hierarchical representation is concerned, nearly half
(49.9) of all servicewomen in NATO are concentrated in the enlisted
ranks, 36% in the NCO's and only 15 % belong to the officers'
Unlike their distribution in terms of occupational areas, the
representation of women throughout the various ranks is much
more balanced when compared to the overall personnel rank structure:
they are only slightly underrepresented in the officer's and
NCO's ranks and over represented among the enlisted. (figure
However, if we look at relative percentages within each hierarchical
category a different reality is highlighted. Due to their very
low absolute numbers, women have an extremely limited representation
in the various ranks. Contradicting the widespread idea that
women's representation is particularly limited in the officer's
ranks if compared to the enlisted force, data shows that there
is not a large gap between those categories: if we exclude from
the analysis those countries that impose rank restrictions (Poland,
Hungary, The Czech Republic, Greece and Turkey) women represent
in average 11.1% of the officers and 12.6% of the enlisted personnel
in NATO forces.
There are relevant differences between countries (Figure 6).
Two patterns emerge: a first one referring to roughly half of
the countries consists of a lower representation of women in
the officer's ranks and a higher representation among enlisted
personnel if compared to women's overall percentage in the force.
Here, the percentage of enlisted women more than doubles their
weight in the officer's categories (Belgium, Denmark, Netherlands,
Luxembourg, Portugal, and Spain).
A second pattern refers to countries where there is a more
balanced representation between categories and even higher percentages
of women in the officer's category than in the enlisted ranks.
Belonging to this group, Norway is a clear 'deviant' case in
that women officers constitute 5,3% of the officer's category,
only 1,5% of enlisted and 3.2% of the total force. In Canada,
the UK and the US women are represented in the officer's category
in slightly higher percentages than in the overall structure;
the same happens in Greece and the Czech Republic, which in
this case may result from existing rank restrictions, since
women are excluded from the enlisted category.
A synthetic index of integration
Despite the lack of information for some of the countries,
the available data - both in terms of precise statistics and
more qualitative analysis of policies - allows the construction
of a global integration index. Considering the macro-analytical
level of this analysis, it is more accurate to talk of an index
of 'formal' integration since the available indicators refer
mainly to global statistics and formal mechanisms and rules.
Different indicators have been chosen to build the index, including
organisational structure and organisational policy indicators
Structural variables refer to the overall representation of
women in active duty forces (1), occupational sex segregation
(3) and rank distribution (5). These indicators are usually
considered of major importance to determine the extent of women's
roles in the military. The impact of related policies is also
captured through the inclusion of indicators pertaining to the
existence of segregation practices (6), and presence or absence
of formal limitations in occupational (2) and hierarchical terms
(4). Hence, beyond the question of relative numbers, other dimensions
of the integration process are included that concern respectively
the structure of opportunities and power distribution. This
is why, together, these indicators contribute with more than
70% to the indexes' overall weight.
Finally, the index includes two additional indicators relative
to existing programs or policies aimed at confronting erosion
factors, such as those derived from the difficult conciliation
between family and a military occupation (6) or sexual harassment
and gender equity monitoring (7). Since it has been recognized
that these factors have a strong impact over integration processes,
attention given by policy-makers to 'quality-of-life' areas
should be taken as important elements for the qualification
of women's presence in the military (6).
It is important to note that this index is not supposed to
reveal absolute positions, from total exclusion to complete
inclusion. While having a low score means almost total exclusion,
having the highest score by no means signifies full integration;
a concrete score merely indicates a relative position between
countries. The country with the highest score is just the one
that has reached the highest level of gender integration within
the selected set of countries. Figure 7 presents a graphic representation
of the countries' scores.
Explaining gender incorporation patterns: a tentative test of
Time effects and relative numbers
- A commonly stated belief is that time has a positive impact
on both quantitative and qualitative dimensions of women's
military integration. According to this perspective, unequal
gender representation is not necessarily the result of discriminatory
practice or 'glass-ceiling' effects, but mainly a consequence
of the limited amount of time women have served in certain
circumstances. From here follows the pretension that time
constitutes a major factor in eliminating discriminations
and bringing nations with lower levels of integration to reach
the more progressive countries. A link is thus established
between time and relative numbers: in general, countries with
a longer experience of female recruitment should also be those
where percentages of women are higher.
- Interestingly though, empirical results do not confirm this
general hypothesis, or at least seem to support it only partially
and for a specific cluster of countries. Data shows a very
low correlation between time and relative numbers (R²
= 0 .26; Sig., 0.05). A graphic representation helps to visualise
this (Figure 8). If there was a high correlation, countries
should cluster along a diagonal linking the upper left and
lower right side of the graphic. This does not seem to happen.
If, besides, we eliminate the extreme cases, where the relation
seems to hold (Italy, Canada and the US), there is a rather
chaotic pattern relating the length of female presence in
the forces and their relative numbers (R² = 0.02).
- If we try to look beyond numbers and strict quantitative
representation, there are reasons to believe results could
be different. The fact that the military is an organisation
where time is usually an important criteria for assessing
statutory positions (rank progression, promotions) makes it
reasonable to admit that time almost 'automatically' affects
various circumstances that define women's positions.
- The relation between the length of female recruitment (Yearstart)
and the index of gender integration (IWMI) shows, in effect,
a significant, if moderate, negative statistical correlation
(R² = 0.41; Sig.- 0.01): the earlier countries started
to recruit women the higher their score is in the integration
index. Figure 9 portrays countries' scores along both dimensions.
- Partial conclusion:
- In sum, results show that time is correlated to some extent
to the degree of women's integration, but this does not hold
true for those countries that score low in integration. Against
common beliefs, data supports the idea that time, by and in
itself, does not automatically foster gender integration and
especially does not have a relevant impact on representation.
Military structure and personnel accession policies
- Drawing on previous theoretical assumptions (Segal, 1995,
2000) and empirical findings (Segal, 1999; Haltiner, 1998a,
1998b), this research assumes the hypothesis that military
variables are major determinants of women's military roles.
Characteristics of force structure are supposed to condition
both representation and other more qualitative variables regarding
the process of gender integration in the Armed Forces (e.g.
- In order to test this general hypothesis (or better, to
check its prevalence for recent years and for the NATO universe),
I have selected, as independent variable, one indicator pertaining
to the organizational format of the military and commonly
used as criteria to define the mass-army character of a force:
one related to personnel accession policies - the conscript
ratio (CR - percentage of conscripts in total active force).
- Results show that there is a negative correlation between
the weight of conscripts in the total active force and the
representation of women in the ranks (women ratio - WR) (R²
= 0.518; Sig. - 0.01). This confirms general research findings
in this area that relate the percentage of women and the organisational
format: the higher the mass army format of a force, the lower
is the WR (figure 10).
- Figure 10 shows that the representation of women is higher
in countries that have voluntary systems of military service
or consider transition from conscript to all-volunteer forces
and face actual or potential recruitment shortages. Inversely,
countries based on conscript military systems and no recruitment
difficulties tend to have the lowest representation of women.
- If we relate the CR with the integration index, results
still point to a significant negative statistical correlation
between both variables: the higher the conscript ratio, the
lower the level of integration. However, this is a moderate
correlation if compared to the previous one (R² = 0.45;
- Partial conclusion:
- We may conclude that organisational format (measured by
the CR) may be a good predictor of women's representation
but does not seem to have the same influence on overall integration.
At the light of this data, we may admit the hypothesis that
other external variables will probably help to better explain
the 'quality' of the process (the level of integration).
The impact of social- economic and political factors: a selective
- A basic hypothesis that has been proposed concerning the
influence of social-economic factors over women's military
participation is that the higher the levels of women's social
and political participation and the more receptive to gender
equality a society is, the less asymmetric should the situation
of women in the military be.
- Two sub-hypothesis were tested:
- Sub-hypothesis 1: 'the greater the percentage
of women in the labor force, the larger will be their representation
in the armed forces',
Sub-hypothesis 2: 'the more gender-egalitarian a
society and the more 'qualified' women's presence in the system,
the more 'inclusive' will be women's military participation'
.(In the absence of serious threats in terms of national and
international security, the degree of gender inclusiveness
and women's 'qualified presence' (Phillips, Jonasdottir, 1990)
in the social and political spheres would affect the degree
of gender inclusiveness in the military).
- a) Considering the impact of women's presence in the
labor force (sub-hypothesis 1):
- Against theoretical expectations, results show that,
for this particular universe, 'women's economic activity
rate' (7) does not affect
their numerical representation in the Armed Forces (R²
= 0.03) nor does it affect integration (IWMI) (R²
= 0.144) (8).
- This result seems to indicate that women's 'simple presence'
or attendance of the system does not seem to be sufficient
to influence their representation or integration levels
within the military.
- b) Considering women's 'qualified presence' in the social
and political spheres
- Among the various possible indicators of women's 'qualified'
or 'controlling' presence I have opted for testing a composite
measure of gender inequality: UNDP's 'Gender Empowerment
Measure'. This measure combines participation and decision-making
indicators in the economic and political spheres as well
as an estimate of power over economic resources (UNDP,
2001: 244-245) (9).
- Here again, data shows no significant relation between
gender inequality (measured by GEM) and relative numbers
of women in the military. However, unlike what happened
with activity rates, there is a significant correlation
between this participation indicator and gender integration
in the military: the higher the GEM (and thus the more
'qualified' the social and political presence of women),
the higher is the integration index (R² = 0.50; Sig.
= 0.01). If, additionally, one of the most deviant cases
is eliminated from the analysis (Germany), the correlation
becomes clearly the higher obtained till now (R²
= 0.724, Sig. = 0.01) (Figure 13).
- Partial conclusion:
- This reinforces an underlying distinction in this research:
that between quantitative indicators or the 'simple presence'
of women and their 'qualified' presence, that is the actual
access to a wider set of opportunities and exercise of power.
These two dimensions are frequently collapsed and it is not
always clear whether scholars refer to relative numbers alone
or to broader 'inclusiveness' dimensions. Additionally, external
variables related to the 'qualification' of women's social
and political participation - more than their economic participation
alone - have a crucial impact on the second dimension ('quality').
If we consider the previous results, a few conclusions can
- Time is not a good predictor of women's representation in
the military: a longer presence of women in the ranks does
not imply a consistent increase in their relative numbers.
Although time is positively correlated to some extent to the
overall degree of women's integration, this only seems be
true for the group of countries that score higher in integration.
Against common beliefs, data supports the idea that time,
by and in itself, does not automatically foster gender integration,
or contribute to eliminate existing discriminations in terms
of occupational or hierarchical de-segregation.
- As far as military related factors are concerned, evidence
shows that, as expected, a specific military factor - the
'organisational format' measured by the 'Conscript Ratio'
- has a significant relation with representation. This means
that the more a force relies on volunteer personnel, the higher
is the percentage of women. Inversely, the closer a military
is to the mass-army format the lower is female representation.
- In opposition to what happens with organisational format,
some 'external' variables related to the social-economic and
political structures seem to have a negligible influence over
numerical representation of women in the military but a significant
impact on their global integration. Still, here one must distinguish
between those factors that concern women's 'simple presence'
in the system, which are not responsible for relevant variation
in the qualitative status of military women and those referring
to women's 'qualified' presence, which instead seem to influence
gender integration to a reasonable extent.
Taken together results suggest that changes in the organisational
format of the military may foster change in the human landscape
of the Armed Forces, but some external variables, specifically
those concerning women's 'qualified' presence in the social
and political realms seem to have a much stronger impact than
sometimes expected on gender integration.
Even if extrapolating somehow beyond available results, I would
hypothesise that change towards greater gender equality in the
Armed Forces will not occur automatically as a consequence of
time or the increase in relative numbers. It will probably depend
much more on the existence of specific policies and the extent
to which external variables, such as women's 'controlling' presence
in society at large, will determine policy orientations and
decision-making processes within the Armed Forces.
paper presents partial results of an ongoing research leading
to the author's PhD thesis. The investigation has been possible
due to the organizational support of the Committee on Women
in the NATO Forces and the financial support of the Portuguese
Ministry of Science (FCT/MCT).
Iceland, that has no Armed Forces.
complete overview of the various factors and associated hypothesis
can be found in Carreiras:2002.
Island doesn't have Armed Forces and Italy started female
recruitment in 2000.
on this variable has been particularly difficult to obtain.
Paradoxical as it may seem if we consider the bureaucratic
nature of the military, many countries do not have systematized
information on the functions that women occupy, and thus national
delegates have failed to provide it in the survey. Methodological
problems related to the complex issue of classifying occupations
also exist that harm the accuracy of comparisons. Nevertheless,
after multiple attempts to obtain this information through
additional personnel contacts and by using multiple sources,
it has been possible to collect data regarding thirteen of
the seventeen NATO nations that employed women in their forces
in 2000 (Island and Italy excluded). Unfortunately it has
not been possible to obtain information regarding percentage
of women within the various occupational areas.
if in some cases it could be argued that the actual over-regulation
of procedures regarding for instance sexual harassment does
not necessarily facilitate integration and may, instead, have
negative collateral effects, we assume that the absence of
institutional channels and rules to identify and solve those
problems does not eliminate them or their damaging consequences
for the overall process of gender integration.
For this indicator I have used data for 1999 (the most recent
available) from the United Nations Development Report (UNDP,
we use a different source, data for 1999 from the UN (Report
of World Development) on female activity rates, these results
are even more accentuated; that is, there is no relation between
female activity rate and the two indicators of women's military
presence I have been using. (Act.rate/Percent women: R²
= 0.03; Act. Rate/Index of integration: R² =.144)
is a combined index measuring gender inequality in three basic
dimensions of empowerment - economic participation and decision-making,
political participation and decision-making and power over
economic resources. It includes the following indicators:
'seats in parliament held by women', female legislators, senior
officials and managers', 'female professional and technical
workers' and ' ratio of estimated female to male earned income'
(cf. UNDP, 2001: 244-245). It varies between 0 and 1.