Killings of women and girls
by their intimate partner
or other family members
Global estimates 2020
3
DATA
MATTERS
11/2021
Some 47,000 women and girls worldwide were killed by their
inmate partners or other family members in 2020. This means
that, on average, a woman or girl is killed by someone in her own
family every 11 minutes. Where trends can be calculated, they
show that the magnitude of such gender-related killings remains
largely unchanged, however, with only marginal increases and
decreases over the past decade.
2 DATA MATTERS 3 11/2021
Acknowledgements
This report was prepared by the Research and Trend Analysis Branch, United Nations
Oce on Drugs and Crime (UNODC).
General coordination and content overview:
Angela Me (Chief, Research and Trend Analysis Branch)
Enrico Bisogno (Chief, Data Development and Dissemination Section)
Analysis and drafting:
Maurice Dunaiski
Andrada-Maria Filip
Data and estimates production:
Camelia Abdelgelil
Antonela Guberac
Tamara Hoefer
Natalia Ivanova
Francesca Rosa
Fatma Usheva
Editing:
Jonathan Gibbons
Cover design and production:
brama Studio
Suzanne Kunnen
Comments:
We thank our colleagues Sven Pfeiffer and Claudia Baroni in the Justice Section of
UNODC for their helpful comments.
Disclaimers
This report has not been formally edited. The contents of this publication can in no way
reect the views or policies of UNODC or contributory organizations, nor do they imply
any endorsement. The designations employed and the presentation of material in this
publication do not imply the expression of any opinion on the part of UNODC concerning
the legal status of any country, territory or city or its authorities, or concerning the delim-
itation of its frontiers or boundaries. This publication may be reproduced in whole or in
part and in any form for educational or non-prot purposes without special permission
from the copyright holder, provided acknowledgement of the source is made. UNODC
would appreciate a copy of any publication that uses this publication as a source.
Contact: UNODC, DDDS: [email protected]
Cover photography: © Heidemarie Pleschko
Copyright © 2021, United Nations Oce on Drugs and Crime
3Gender-related killings of women and girls in the private sphere
Killings of women and girls by their intimate partneror other family members
KEY FINDINGS
• Some 47,000 women and girls worldwide were killed by their intimate partners or other family mem-
bers in 2020. This means that, on average, a woman or girl is killed by someone in her own family every
11 minutes.
• Women and girls in all regions across the world are affected by this type of gender-based violence.
With an estimated 18,600 victims, Asia is the region with the highest number of victims in absolute
terms, while Africa is the region with the highest level relative to the size of its female population.
• Accounting for 80 per cent of the total, the vast majority of all homicide victims worldwide in 2020
were men or boys, but the sex composition of homicide victims differs markedly depending on the
context. Women and girls account for only one tenth of all homicide victims perpetrated in the pub-
lic sphere, yet they bear a disproportionate burden of lethal violence perpetrated in the home: in 58 per
cent of all killings perpetrated by intimate partners or other family members, the victim was a woman
or girl.
• National trends in gender-related killings of women and girls are often volatile. On average, Europe
has witnessed a 13 per cent decline in gender-related killings of women and girls in the private sphere
over the past decade, while the Americas has seen an increase of 9 per cent, although these averages
mask contrasting trends between the countries and subregions within those two regions. Due to the
limited availability of data, it is not yet possible to identify time trends in gender-related killings in other
regions.
• Lethal violence experienced by women and girls in the private sphere appears to be a more intractable
problem than killings of women and girls outside the home. Eastern Europe has witnessed a reduction
of 47 per cent in female homicides perpetrated outside the home over the past decade, while female
homicides perpetrated by intimate partners or other family members have seen a reduction of 15 per
cent over the same period. Although less marked, similar trends have also been recorded in Western
Europe and South America.
• Global data on the impact of COVID-19 connement measures on gender-related killings of women and
girls remain patchy and inconclusive. The average annual number of these killings showed an increase
of 11 per cent from 2019 to 2020 in Western Europe and of 5 per cent in Southern Europe. No change
was detectable in Northern Europe and there was a slight decrease in Eastern Europe. By comparison,
in Northern America the number of gender-related killings of women and girls increased by 8 per cent,
in Central America by 3 per cent and in South America by 5 per cent. These increases were of a similar
magnitude to previous annual changes recorded in the past decade.
• At the national level, monthly data from a limited number of countries covering the period October
2019 to December 2020 show high variability in trends across countries, but suggest that, overall,
female intimate partner/family-related homicides remained relatively unaffected by the lockdowns in
those countries.
4 DATA MATTERS 3 11/2021
Key ndings
Global estimates
Women and girls in all regions are affected by gender-based killings. While Asia is
the region with the largest absolute number of killings, Africa is the region with the
highest level of violence relative to the size of its female population.
2020
2020
Total female intimate partner/family-related homicides
Rates of female intimate partner/family-related homicide per 100,000 female population
Asia Africa Americas Europe Oceania
1.2 per 100,000 female
population
0.8 2.7 1.4 0.7 1.6
18,600 18,100 7,300 2,600 300
47,000
worldwide
Asia Africa Americas Europe Oceania
5Gender-related killings of women and girls in the private sphere
Key ndings
2020
Gender dimension of homicide
Homicides within the private sphere affect both sexes, but women and girls bear
the largest burden of lethal violence perpetrated within the home, accounting for
approximately 6 out of every 10 homicide victims killed by intimate partners or other
family members.
female
19%
male
81%
Total homicides
Intimate partner/
family-related homicides
2020
Non-Intimate partner/
family-related homicides
2020
female
58%
male
42%
female
10%
male
90%
6 DATA MATTERS 3 11/2021
Key ndings
20202010
2020
2019
Trends in female intimate partner/family-related
homicide
13%
9%
10
Years
female intimate partner/family-related
homicide over 10 years
decrease in Europe
in Eastern Europe, female intimate partner/family-
related homicides decreased significantly less than
homicides perpetrated outside the domestic sphere
number of homicides of women and girls by intimate partners
or other family members increased from 2019 to 2020 in
almost all subregions where trends could be estimated
no change in Northern Europe and slight decrease in Eastern Europe
3. On average, Europe has
witnessed a 9 per cent decline in
gender-related killings in the
private sphere over the past
decade, while the Americas saw
an increase by 8 per cent, albeit
with significant differences
between countries within the
regions. Due to limited data
availability, it is not yet possible
to describe time trends in gen-
der-related killings in other world
regions.
5. Lethal violence experienced
by women in the private sphere
appears to be a more intractable
problem than lethal violence outsi-
de the home. For example, Europe
has witnessed a remarkable 45 per
cent reduction in female homicides
outside the home over the last
decade, while female homicides
perpetrated by intimate partners or
other family members have seen a
more modest reduction of 9 per
cent over the same period.
6. Global data on the impact
of COVID-19 confinement measu-
res on gender-related killings
remain patchy and inconclusive.
Monthly data on a limited number
of countries - 15 in different
regions - show that female homi-
cides continued at approximately
the same pace throughout the
various waves of COVID-related
mobility restrictions.
Western
Europe
Northern
America
Southern
Europe
Central
America
South
America
+8%
increase in the Americas
+11%+5%+5%+3%
7Gender-related killings of women and girls in the private sphere
Key ndings
47,000
worldwide
Total number of women killed by intimate
partners or other family members
In 2020, approximately 47,000 women and girls were killed worldwide by their
intimate partners or other family members. This means that, on average, a
woman or girl is killed by someone in her own family every 11 minutes.
3. On average, Europe has
witnessed a 9 per cent decline in
gender-related killings in the
private sphere over the past
decade, while the Americas saw
an increase by 8 per cent, albeit
with significant differences
between countries within the
regions. Due to limited data
availability, it is not yet possible
to describe time trends in gen-
der-related killings in other world
regions.
5. Lethal violence experienced
by women in the private sphere
appears to be a more intractable
problem than lethal violence outsi-
de the home. For example, Europe
has witnessed a remarkable 45 per
cent reduction in female homicides
outside the home over the last
decade, while female homicides
perpetrated by intimate partners or
other family members have seen a
more modest reduction of 9 per
cent over the same period.
6. Global data on the impact
of COVID-19 confinement measu-
res on gender-related killings
remain patchy and inconclusive.
Monthly data on a limited number
of countries - 15 in different
regions - show that female homi-
cides continued at approximately
the same pace throughout the
various waves of COVID-related
mobility restrictions.
1 woman or girl is killed by
someone in her own family
every
11
minutes
Killings of women and girls by their intimate partner or other family members
8
1. WHY FOCUS ON KILLINGS OF WOMEN AND GIRLS PERPETRATED BY INTIMATE PARTNERS OR OTHER
FAMILY MEMBERS? ....................................................................................................................................... 9
2. EXTENT OF FEMALE INTIMATE PARTNER/FAMILY-RELATED HOMICIDE IN ABSOLUTE NUMBERS ............. 13
3. EXTENT OF FEMALE INTIMATE PARTNER/FAMILY-RELATED HOMICIDE IN RATES ..................................... 14
4. TRENDS IN FEMALE INTIMATE PARTNER/FAMILY-RELATED HOMICIDE ..................................................... 14
5. HOMICIDE PERPETRATED IN THE PRIVATE SPHERE HAS A DISPROPORTIONATE IMPACT ON WOMEN ..... 18
6. BURDEN OF INTIMATE PARTNER HOMICIDE BORNE BY WOMEN AND GIRLS ............................................ 19
7. WHAT IMPACT HAS THE COVID-19 PANDEMIC HAD ON GENDER-RELATED KILLINGS? ............................. 20
8. POLICY IMPLICATIONS ................................................................................................................................ 25
9. STATISTICAL ANNEX .................................................................................................................................... 28
10. METHODOLOGICAL ANNEX ......................................................................................................................... 31
Killings of women and girls by their intimate partner or other family members
9
The killing of women and girls at the hands of intimate partners or other family members – people whom
they would normally be expected to trust – represents one of the most extreme manifestations of gender-
based violence. Such killings are often the culmination of prior experiences of gender-based violence, which
can include psychological, sexual and physical abuse.
1
Female homicide perpetrated by intimate partners or other family members is not the only form of gender-
related killing of women and girls. This type of homicide can also be perpetrated outside the private sphere,
for example, in the case of intentional homicides of female sex workers, killings of women related to various
forms of trafficking and organized crime, and killings of women accused of witchcraft. Several types of
female homicide are committed on the grounds of gender-related motivations
2
but there is not yet a
standardized approach to define and count all forms of gender-related killings of women and girls. When
available, statistical data on this type of homicide are often based on national definitions, practices or
specific legislation that are not comparable at the global level. To address this issue, the United Nations
system has held a series of consultations aimed at developing a new statistical approach to improving data
on gender-related killings of women and girls (see box below).
This research brief is focused on female victims of intentional homicide perpetrated by intimate partners
or other family members. This is because – based on available evidence – intimate partner/family-related
homicide accounts for the largest proportion of gender-related killings or “femicides”.
3
Furthermore, data
on this type of homicide are more readily available and comparable than data on gender-related killings
that take place outside the family sphere. Internationally comparable data on female intimate
partner/family-related homicide remain patchy, but are available for a large enough number of countries
to enable the production of global and regional estimates,
4
thanks also to the implementation of the
framework provided in the International Classification of Crime for Statistical Purposes (ICCS).
5
Although
still affected by limitations caused by poor coverage, in particular in Africa and Asia, these data provide
useful information on the magnitude of the problem and, in some cases, information on trends. Data on
gender-related killings perpetrated outside the family sphere are more limited and, to date, do not meet
the same level of standardization and geographical coverage.
1
UNODC, Global Study on Homicide 2019, booklet 5, Gender-related Killing of Women and Girls (United Nations publication, 2019).
2
General recommendation No. 35 on gender-based violence against women, updating general recommendation No. 19, para. 19
(CEDAW/C/GC/35).
3
UNODC, Global Study on Homicide 2019, booklet 5, Gender-related Killing of Women and Girls (United Nations publication, 2019), p.25.
4
During the period 2010–2020, 95 countries made available at least once data on intentional homicides of female victims perpetrated by
their intimate partner or other family members. These data were submitted to UNODC by Member States through the annual United
Nations Surveys on Crime Trends and the Operations of Criminal Justice Systems (UN-CTS). Where applicable, UN-CTS data are
supplemented with external data sources. Regional groups are based on the UN M49 classification scheme.
5
UNODC, International Classification of Crime for Statistical Purposes (ICCS) (Vienna, 2017).
Killings of women and girls by their intimate partner or other family members
10
STATISTICAL FRAMEWORK FOR MEASURING GENDER-RELATED KILLINGS OF WOMEN AND GIRLS
(ALSO REFERRED TO AS “FEMICIDES/FEMINICIDES”)
Different approaches to and definitions of gender-related killings of women and girls (also referred to
as “femicide and/or feminicide”) exist at the national and international levels.
6
For example, the 2015
report of the United Nations Secretary General on Action against gender-related killing of women and
girls referred to this concept as “the intentional murder of women and girls because of their gender”.
7
From a conceptual point of view, such killings differ from other types of intentional killings of women
and girls because they are based on gender-related motivations, either directly or indirectly.
In order to harmonize the existing statistical approaches for measuring gender-related killings of women
and girls, in 2019 the United Nations Statistical Commission
8
asked UNODC to develop a “statistical
framework on gender-sensitive crime statistics, with a focus on the gender-related killing of women and
girls (femicide/feminicide)”. UNODC, together with the United Nations Entity for Gender Equality and
the Empowerment of Women (UN Women), undertook a thorough expert-level consultation process in
order to develop a global statistical framework for measuring gender-related killings of women and girls,
aligned with the International Classification of Crime for Statistical Purposes (ICCS); this box briefly
describes the proposed conceptualization of the statistical framework.
9
In general terms, gender-related killings of women and girls are committed on the grounds of “gender-
related factors such as the ideology of men’s entitlement and privilege over women, social norms
regarding masculinity, and the need to assert male control or power, enforce gender roles, or prevent,
discourage or punish what is considered to be unacceptable female behavior”.
10
Gender-related killings
can take place in a wide range of situations within both the private and public spheres, and within
different contexts of victim–perpetrator relationship. In principle, what defines a gender-related killing
is not a specific relationship or type of perpetrator but the fact that it can be rooted in the gender-related
motivations described above.
Such killings have a clear conceptual meaning, but they are challenging to operationalize in statistical
terms as this requires an understanding of which behaviours, modus operandi and circumstances
associated with a killing are actually gender-related.
11
The following set of non-exhaustive standard
characteristics related to the modus operandi or context of an intentional homicide has been identified
as indicative of gender-related motivations in killings of women and girls.
12
6
For further reference, see United Nations, “Report of the Special rapporteur on violence against women, its causes and consequences,
Rashida Manjoo”, paras. 14-18 (A/HRC/20/16); UNODC, Global Study on Homicide 2019, booklet 5, Gender-related Killing of Women and
Girls (United Nations publication, 2019).
7
United Nations, “Action against gender-related killing of women and girls”, Report of the Secretary-General (2015), para. 2
(A/70/93).
8
(E/CN.3/2019/19).
9
UNODC and UN Women, “Statistical framework for measuring the gender-related killings of women and girls (also
referred to as “femicides/feminicides”)”, forthcoming.
10
General recommendation No. 35 on gender-based violence against women, updating general recommendation No. 19,
para. 19 (CEDAW/C/GC/35).
11
UNODC, Global Study on Homicide 2019, booklet 5, Gender-related Killing of Women and Girls (United Nations
publication, 2019.
12
For further information, see UNODC and UN Women, “Statistical framework for measuring the gender-related killing of
women and girls (also referred to as “femicide/feminicide”)” (forthcoming).
Killings of women and girls by their intimate partner or other family members
11
Conceptual framework for measuring gender-related killings of women and girls
(femicide/feminicide)
Characteristics indicative of gender-related motivations of killings of women and girls
(femicide/feminicide)
In principle, these variables could be used to identify gender-related killings among intentional
homicides of women and girls, but they are not yet mainstreamed in national statistical systems and
data on gender-related killings cannot be readily collated at the national and international levels purely
on the basis of these characteristics. Until national statistical systems fully adopt these standard
parameters, the first building block in the construction of data on gender-related killings can be
approximated on the basis of women and girls killed by partners or other family members. While the
victim’s relationship with a perpetrator who is a partner or other family member is not necessarily
indicative of specific gender motivations by itself, evidence indicates that this type of homicide is
overwhelmingly related to gender roles. Data on women and girls killed by partners and family members
are also more readily available because information on the relationship between the homicide
perpetrator and victim is already recorded in a significant number of national systems. Being aware that
an approximation is made on the grounds of data availability and statistical feasibility, data on women
killed by intimate partners and on women killed by other family members can be used as the first
building block for counting gender-related killings.
Killings of women and girls by their intimate partner or other family members
12
The other building block to count gender-related killings lies outside the family sphere, where the eight
“modus operandi” or contexts remain the only parameters that can operationalize the gender-relevant
subset of homicide.
Statistical framework for measuring gender-related killing of women and girls
The aim of the new statistical framework is also to provide operational guidance to national statistical
systems; several institutions can produce data on gender-related killings and the harmonization of
practices across national institutions fosters comparative analyses of data. For example, the analysis of
relevant data across stages of the criminal justice process – from cases and prosecutions to convictions
– can provide important information about state response in relation to gender-related killings of
women and girls. Futhermore, the framework is aimed at providing guidance on the collection of
additional information (disaggregating variables) for to conducting policy-relevant analyses to identify
drivers and enablers of gender-related killings and factors that could help prevent them.
Killings of women and girls by their intimate partner or other family members
13
In 2020, it is estimated that – out of a total of 81,00 female victims of homicide – 47,000 women (41,500–
52,600)
13
and girls worldwide were killed by their intimate partners or other family members (including
fathers, mothers, uncles and brothers).
14
With an estimated 18,600 victims (15,100–21,000), Asia
accounted for the largest number of female intimate partner/family-related killings in 2020. This was
followed by Africa, with 18,100 (16,500–20,800), and the Americas, with 7,300 (7,000–7,600), Europe
accounted for 2,600 (2,500–2,700) female intimate partner/family-related killings in 2020 and Oceania
recorded the lowest number, with 300 (250–350) (figure1).
15
It is important to note, however, that these
absolute numbers are also influenced by the differences in the size of the populations of the world regions.
With an estimated population of 4.6 billion in 2020, Asia is by far the most populous region, so it is not
surprising that the continent accounts for the largest share of female intimate partner/family-related
homicides globally.
Estimated number of female victims of intimate partner/family-related homicide, by region
(2020)
Source: UNODC estimates based on UNODC Homicide dataset.
Note: Error bars represent estimation error due to imputation of missing values at the country level. The bands do not
represent confidence intervals and should not be interpreted as such. The estimates are based on the latest available data,
which remain limited, in particular in Africa and Asia. The size of the error bars provides an indication of the level of
uncertainty that the estimates carry both at the regional level.
13
The global and regional estimates are produced on the basis of available national data and through a statistical model imputing missing
values at the country level. They are complemented by an interval of uncertainty that accounts for estimation error due to the imputation
of missing values at the country level. The bands do not represent confidence intervals and should not be interpreted as such. For more
information, see the methodological annex to this document.
14
The 2020 estimate of 47,000 victims is not directly comparable to the 2017 figure (50,000 victims) published in the Global Study on
Homicide 2019 due to an improvement in the coverage of national data and refinements in the statistical model. The different global
estimates for 2017 and 2020 should not be interpreted as evidence of a decrease in the global number of gender-related killings in the
private sphere over this time span.
15
The difference of 100 killings between the estimated global total and the sum of the regional estimates is due to rounding.
18,100
7,300
18,600
2,600
300
- 5,000 10,000 15,000 20,000 25,000
Africa
Americas
Asia
Europe
Oceania
Estimated number of female intimate partner/family-related homicides
(2020)
Killings of women and girls by their intimate partner or other family members
14
From a comparative perspective, it is more informative to look at the number of female victims of intimate
partner/family-related homicide in different regions per 100,000 female population than in absolute
numbers, as the former takes into account differences in population size and growth. Victimization rates
are also better than absolute numbers at capturing women’s actual risk of being killed by someone in their
family.
In 2020, it is estimated that out of every 100,000 women and girls worldwide, 1.2 were killed by their
intimate partners or other family members.
16
While Asia accounts for the largest absolute number of
female victims of intimate partner/family-related homicide, it is in Africa that women and girls appear to
be most at risk of being killed by their intimate partners or other family members. In 2020, the female
intimate partner/family-related homicide rate in Africa was estimated at 2.7 per 100,000 female
population, compared with 1.6 in Oceania, 1.4 in the Americas, 0.8 in Asia and 0.7 in Europe.
The variability in the rate of female intimate partner/family-related homicide across the five world regions
was significantly lower in 2020 than the variability in the total homicide rate, which ranged from roughly
2.5 in Europe to almost 15.0 in the Americas.
17
Thus, while the overall level of lethal violence varies
significantly across the regions, the variability is less pronounced in the case of female killings within
domestic contexts.
In those regions where data availability allows for an investigation of trends over time (the Americas and
Europe), substantial differences can be observed in trends in the annual number of female intimate
partner/family-related homicides. Both in Europe and the Americas, data availability and national trends
vary, with some countries displaying increases over certain time periods, some displaying decreases and
some a level of year-on-year variability or irregular data availability that make it impossible to discern clear
trends.
18
However, what does seem to be common to many national trends is that they show relative
stability when longer periods of time are considered, rather than year-on-year changes.
Regional estimates level out different national trends but, overall, Europe has seen a modest reduction (13
per cent) in the total number of female victims of intimate partner/family-related homicide over the past
decade, while there has been a similar level of increase in the Americas (9 per cent) over the same period,
albeit at a very different absolute level (figure 2).
16
Due to improvements in data coverage and statistical modelling, the estimated 2020 global rate of 1.2 female intimate partner/family-
related homicides per 100,000 women is not directly comparable to the 2017 rate of 1.3 rate published in the Global Study on Homicide
2019 and should not be interpreted as a decrease in the risk of such killings at the global level in recent years.
17
UNODC Homicide data, preliminary estimates for 2020 (forthcoming).
18
For the variability in Europe, see figure 17 in the annex to this document; for the variability in the Americas, see figure 18.
Killings of women and girls by their intimate partner or other family members
15
Trends in the total number of female victims of intimate partner/family-related homicide in
the Americas and Europe (2010–2020)
Source: UNODC estimates based on UNODC Homicide dataset.
Note: The dotted lines represent ranges around the estimate and show the extent to which the regional estimate for a
specific year relies on reported versus imputed country-level series of intimate partner/family-related homicide. The ranges
do not represent confidence intervals and should not be interpreted as such. For more information, see the accompanying
methodological annex.
The modest reduction in the total number of female victims of intimate partner/family-related homicide
in Europe is the result of national trends that fluctuate around a small overall decline in several countries
in Eastern Europe, including Belarus and Romania, and a few countries in Western Europe. In several
populous European countries such as Germany, France, Italy, Spain and the United Kingdom there has been
variability around stable levels of gender-related killings in the private sphere in the last few years (figure
19).
The overall trend in female intimate partner/family-related homicides witnessed in the Americas over the
last decade is the result of divergent trends experienced at the country level. For example, among the most
populous countries in the region, the United States and Brazil have recorded a slight decrease in recent
years, while there has been an upward trend in Mexico (figure 20).
19
By contrast, less populous countries,
but with higher rates of female intimate partner/family-related homicide, such as El Salvador and Honduras,
have recorded some reductions since 2015 (figure 20). Overall, the Americas are characterized by
substantial heterogeneity at the country level when it comes to levels of and trends in intimate
partner/family-related homicides and other gender-based killings.
The average trends in Europe and the Americas mask differences and commonalities between subregions.
While in Western, Southern and Northern Europe, the number of female homicides within the private
sphere is consistently higher than in the public sphere, trends in Eastern Europe (figure 3) and the other
subregions of Europe (figure 4) suggest that female intimate partner/family-related killings are harder to
eradicate than other types of killings of women and girls. Both subregions have seen a noticeable decrease
in the number of female victims of killings outside the family over the past decade (47 per cent in Eastern
Europe and 21 per cent in Western Europe), whereas the number of female homicides perpetrated by
intimate partners or other family members has remained relatively unchanged. As a result of these
different trends, the share of total female homicide victims that were killed by intimate partners or other
family members has increased substantially, from 34 per cent in 2010 to 45 per cent in 2020 in Eastern
Europe and from 60 per cent to 63 per cent over the same period in the other subregions of Europe.
19
For more information on the situation in Mexico, see Calderón, L. et al. (eds.), Organized Crime and Violence in Mexico: 2021 Special
Report (Justice in Mexico, University of San Diego, October 2021).
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Number of female intimate partner/family
-
related homicides
Americas Europe
Killings of women and girls by their intimate partner or other family members
16
Trends in female homicides in the private and public sphere in Eastern Europe (2010–2020)
Source: UNODC estimates based on UNODC Homicide dataset.
Note: Eastern Europe is defined according to the M49 regional grouping.
Trends in female homicides in the private and public sphere in Western, Northern and
Southern Europe (2010–2020)
Source: UNODC estimates based on UNODC Homicide dataset.
A similar pattern can be observed in South America where, from 2017 to 2020, female homicides
perpetrated outside the private sphere decreased by 20 per cent while they remained relatively stable
within the family (figure 5). These data suggest that factors that triggered the decrease in female non-
intimate partner/family-related homicides in Europe and South America only partially impacted on
intimate partner/family-related homicides. They also suggest that positive contextual developments and
policy interventions that are successful in reducing lethal violence in the public sphere, may not be
sufficient to achieve the same reductions in gender-based killings within the family.
0
10
20
30
40
50
60
70
0
500
1000
1500
2000
2500
3000
3500
4000
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Intimate partner/family
-related share of
female homicides (%)
Number of victims
Intimate partner/family-related share of female homicides
Intimate partner/family-related female homicides
Non-intimate partner/family-related female homicides
0
10
20
30
40
50
60
70
0
200
400
600
800
1000
1200
1400
2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
2020
Intimate partner/family
-related share of
female homicides (%)
Number of victims
Intimate partner/family-related share of female homicides
Intimate partner/family-related female homicides
Non-intimate partner/family-related female homicides
Killings of women and girls by their intimate partner or other family members
17
Trends in female homicides in the private and public sphere in South America (2017–2020)
Source: UNODC estimates based on UNODC Homicide dataset.
Note: South America is defined according to the M49 regional grouping.
Trends in the number of women and girls killed in the private and public sphere are flatter in Central
America (figure 6) and Northern America (figure 7), although the slight increase in non-intimate
partner/family-related killings in Central America does not seem to have affected the stable trend in
intimate partner/family-related killings. This suggests that female intimate partner/family-related killings
tend to remain stable even in contexts where there are dynamics that trigger increases in non-family
violence.
Trends in female homicides in the private and public sphere in Central America (2017–2020)
Source: UNODC estimates based on UNODC Homicide dataset.
Note: Central America refers to Central America and the Caribbean, as defined according to the M49 regional grouping.
0
10
20
30
40
50
60
70
0
1000
2000
3000
4000
5000
6000
2017 2018 2019 2020
Intimate partner/family
-related share of
female homicides (%)
Number of victims
Intimate partner/family-related share of female homicides
Intimate partner/family-related female homicides
Non-intimate partner/family-related female homicides
0
10
20
30
40
50
60
70
0
500
1000
1500
2000
2500
3000
3500
4000
2017 2018 2019 2020
Intimate partner/family
-related share of
female homicides (%)
Number of victims
Intimate partner/family-related share of female homicides
Intimate partner/family-related female homicides
Non-intimate partner/family-related female homicides
Killings of women and girls by their intimate partner or other family members
18
Trends in female homicides in the private and public sphere in Northern America (2016–
2019)
Source: UNODC estimates based on UNODC Homicide dataset.
Note: Northern America is defined according to the M49 regional grouping. Data availability constraints for the year 2020
mean that this year is omitted from the trend analysis for Northern America.
In Africa, Asia and Oceania, it is not possible to identify trends in female intimate partner/family-related
homicide over the past decade, because of the limited availability of sufficiently disaggregated homicide
data covering the period 2010–2020. While the link between trends in female homicide and those
perpetrated by intimate partner or other family members is not always direct (see section 4), the significant
reduction in female homicides in certain populous countries, such as India, Japan, the Philippines and
Turkey, since 2010 may indicate that the situation is slowly improving in Asia. This is substantiated to some
extent by country-level data on gender-related violent deaths from populous countries such as India, where
violent deaths related to dowry payments, accusations of witchcraft and other gender-related factors have
been slowly declining (figure 21).
In several populous countries in Africa, including Morocco, Kenya, South Africa and Uganda, numbers of
total female victims of intentional homicides in the last decade are not indicative of clear trends, but show
that the level of lethal violence perpetrated against women and girls in the region remains high in relative
terms (figure 23). In Oceania, the two largest countries (Australia and New Zealand) have both recorded
little progress in terms of reducing the annual number of female intimate partner/family-related killings
over the past decade (figure 22).
Accounting for 80 per cent of the total, the vast majority of all homicide victims worldwide in 2020 were
men or boys, but the sex composition of homicide victims differs markedly depending on the context.
Women and girls account for only one tenth of all homicide victims perpetrated in the public sphere, yet
they bear a disproportionate burden of lethal violence perpetrated within the home: in 58 per cent of all
killings perpetrated by intimate partners or other family members, the victim was a woman or girl.
Out of all intentional homicides of women and girls in 2020, 58 per cent were perpetrated by intimate
partners or other family members, which suggests that the most dangerous place for women is the home.
By contrast, the share of all male homicides that can be attributed to intimate partners or other family
members amounted to just 10 per cent in 2020, suggesting that men and boys are primarily at risk of being
intentionally killed outside the family sphere.
0
10
20
30
40
50
60
70
0
500
1000
1500
2000
2500
3000
2016 2017 2018 2019
Intimate partner/family
-related share of
female homicides (%)
Number of victims
Intimate partner/family-related share of female homicides
Intimate partner/family-related female homicides
Non-intimate partner/family-related female homicides
Killings of women and girls by their intimate partner or other family members
19
These gender-specific patterns of victimization differ substantially across regions (figure 8). For example,
in regions with lower overall levels of intentional homicide such as Europe and Oceania, the share of male
homicides related to violence in the family sphere is larger, at 17 per cent and 22 per cent respectively,
than in regions with very high levels of homicidal violence such as Africa and the Americas.
Share of intimate partner/family-related homicides among all female and male homicides, by
region (2020)
Source: UNODC estimates based on UNODC Homicide dataset.
Intimate partner homicide represents the ultimate breach of trust within the home. Moreover, this type of
homicide has consequences that go far beyond the direct victims.
20
For example, the surviving offspring
not only lose one parent to homicide but may also lose the other to prison, and they are often obliged to
leave their parental home as a result.
21
Based on available data,
22
homicides of female intimate partners account for an average of roughly 67 per
cent of all female intimate partner/family-related killings,
23
which suggests that killings by current and
former intimate partners account for the largest proportion of all female intimate partner/family-related
killings. This means that women and girls bear an even greater burden in the case of intimate partner
homicide than of intimate partner/family-related killings in general: on average, women and girls account
for approximately 68 per cent of all victims of intimate partner homicide.
24
Moreover, it appears that the burden of intimate partner homicide borne by women and girls has remained
relatively stable in recent years: in 17 countries in various regions, the disparity between the annual
20
Children can also become direct victims of intimate partner violence, for example, in cases of “familicide” or when children are killed in a
deliberate attempt to make the intimate partner suffer. See UNODC, Global Study on Homicide 2019, booklet 6, Killing of Children and
Young Adults (United Nations publication, 2019).
21
UNODC, Global Study on Homicide 2019, booklet 5, Gender-related Killing of Women and Girls (United Nations publication, 2019).
22
Data on female homicides perpetrated exclusively by intimate partners are available from 67 countries across all regions for at least one
year in the period 2015–2020.
23
Average weighted by the number of female intimate partner/family-related killings in each of the 67 countries.
24
Note that the average is weighted by the number of intimate partner/family-related killings (male and female victims) in each of the 67
countries, giving more weight to more populous countries.
72
40
59
51
75
58
9
7
16
17
22
10
0
10
20
30
40
50
60
70
80
Africa Americas Asia Europe Oceania World
Intimate partner/family
-related homicide
share of all hoicides, by sex (%)
Female intimate partner/family-related homicide share of all female
homicides
Male intimate partner/family-related homicide share of all male homicides
Killings of women and girls by their intimate partner or other family members
20
numbers of male and female intimate partner killings did not change significantly over the period 2015–
2020 (figure 9).
25
Trends in male and female intimate partner killings in 17 selected countries (2015–2020)
Source: UNODC Homicide dataset.
Note: Based on data from 17 countries where data was available (Croatia, Germany, Guyana, Iceland, Italy, Latvia,
Lithuania, Mexico, Montenegro, Morocco, Netherlands, Paraguay, Russian Federation, Slovakia, Slovenia, Spain and
Switzerland).
Global data on the impact of COVID-19 confinement measures on gender-related killings of women and
girls remain patchy and inconclusive. The average annual number of these killings showed an increase of
11 per cent from 2019 to 2020 in Western Europe and of 5 per cent in Southern Europe. No change was
detectable in Northern Europe and there was a slight decrease in Eastern Europe. By comparison, in
Northern America the number of gender-related killings of women and girls increased by 8 per cent, in
Central America by 3 per cent and in South America by 5 per cent. These increases were of a similar
magnitude to previous annual changes recorded in the past decade.
At the national level, monthly data from 14 countries in various regions show high variability in trends
across countries but suggest that, overall, female intimate partner/family-related homicides remained
relatively unaffected by the lockdowns in those countries.
25
The slight difference in the female share of intimate partner homicide victims between this group of 17 countries (63 per cent in 2020)
and the female share of intimate partner homicide victims in the larger group of 67 countries referred to earlier (67 per cent in 2020) is
because of the inclusion of different countries as well as the weighting approach applied the larger group of countries.
40
45
50
55
60
65
70
75
80
0
200
400
600
800
1000
1200
1400
1600
2015 2016 2017 2018 2019 2020
Female share of intimate partner homicides (%)
Number of victims
Female share of intimate partner homicides (%)
Female intimate partner homicides
Male intimate partner homicides
Killings of women and girls by their intimate partner or other family members
21
Indexed monthly number of female victims of intimate partner/family-related homicide in
14 countries (October 2019–December 2020)
Source: UNODC, The Impact of COVID-19 on Drugs and Crime. Available at https://www.unodc.org/unodc/en/data-and-
analysis/coronavirus.html.
Note: Based on monthly data from 14 countries (Armenia, Bahamas, Croatia, Ecuador, Greece, Italy, Latvia, Lithuania,
Mexico, Morocco, Myanmar, Oman, Slovenia and Spain).
The broader impact of COVID-19 mobility restrictions on other types of gender-based violence also remains
unclear because of the limited availability of data. Evidence from 11 countries in various regions suggests
that the number of cases of sexual violence (figure 11) and physical assault (figure 12) perpetrated by
intimate partners or other family members reported to authorities declined briefly during the first wave of
confinement measures in early 2020 but reverted to the previous level as countries emerged from the
initial lockdown. This suggests that during the first wave of confinement measures there was likely both a
decrease in the accessibility of women to institutions for reporting incidents of domestic violence and in
the capacity of authorities to record those incidents.
26
26
UNODC Research brief, “What crime and helpline data say about the impact of the COVID-19 pandemic on reported violence against
women and girls” (2020).
0
100
200
300
400
500
600
Oct 19
Nov 19
Dec 19
Jan 20
Feb 20
Mar 20
Apr 20
May 20
Jun 20
Jul 20
Aug 20
Sep 20
Oct 20
Nov 20
Dec 20
Indexed number of monthly female victims
(Baseline Feb 2020 = 100)
● Number of female
intimate
partners/family-
related homicides in
14 countries, relative
to baseline (Feb 2020
= 100)
● Country-level
trends relative to
baseline
(Feb 2020 = 100)
● Confinement
measures
Killings of women and girls by their intimate partner or other family members
22
Indexed monthly number of female victims of sexual violence perpetrated by intimate
partners or other family members, as reported to authorities in 11 countries (October 2019–
March 2021)
Source: UNODC, The Impact of COVID-19 on Drugs and Crime. Available at https://www.unodc.org/unodc/en/data-and-
analysis/coronavirus.html.
Note: Based on monthly data on sexual violence cases from 11 countries (Armenia, Croatia, Latvia, Mongolia, Morocco,
Myanmar, Poland, Slovenia, Spain, Sweden and Switzerland)
Indexed monthly number of female victims of physical assault perpetrated by intimate
partners or other family members, as reported to authorities in 10 countries (October 2019–
March 2021)
Source: UNODC, The Impact of COVID-19 on Drugs and Crime. Available at https://www.unodc.org/unodc/en/data-and-
analysis/coronavirus.html.
-50
0
50
100
150
200
250
300
Oct 19
Nov 19
Dec 19
Jan 20
Feb 20
Mar 20
Apr 20
May 20
Jun 20
Jul 20
Aug 20
Sep 20
Oct 20
Nov 20
Dec 20
Jan 21
Feb 21
Mar 21
Monthly number of victims relative to the
baseline
(Feb 2020 = 100)
*
11 countries: Armenia, Croatia, Latvia, Mongolia, Morocco, Myanmar (Burma), Poland, Slovenia, Spain,
Sweden, Switzerland
● Number of female
victims of sexual
violence by intimate
partners or family
members in 11
countries, relative to
baseline
(Feb 2020 = 100)
● Country-level trends
relative to baseline
(Feb 2020 = 100)
● Confinement
measures
-50
0
50
100
150
200
250
300
Oct 19
Nov 19
Dec 19
Jan 20
Feb 20
Mar 20
Apr 20
May 20
Jun 20
Jul 20
Aug 20
Sep 20
Oct 20
Nov 20
Dec 20
Jan 21
Feb 21
Mar 21
Monthly number of victims relative to the baseline
(Feb 2020 = 100)
**
10 countries: Croatia, Latvia, Lithuania, Macao SAR China, Morocco, Namibia, Poland, Slovenia,
Spain, Switzerland
● Number of victims
of physical assault
by intimate partners
or family members
in 10 countries,
relative to baseline
(Feb 2020 = 100)
● Country-level
trends relative to
baseline
(Feb 2020 = 100)
● Confinement
measures
Killings of women and girls by their intimate partner or other family members
23
Note: Based on monthly data on sexual violence cases from 10 countries (Croatia, Latvia, Lithuania, Macao, China,
Morocco, Namibia, Poland, Slovenia, Spain and Switzerland).
Domestic violence helpline data from several countries in Europe and Latin America also paint a mixed
picture of the impact of COVID-19 confinement measures on the level of violence experienced by women
and girls in the home. Some national helpline data suggest a sharp increase in reported violence against
women in the home. For example, data from the national domestic violence helpline in Italy show a drastic,
four-fold increase
27
in the weekly number of calls reporting a case of violence at the start of the first
lockdown, which started on 9 March 2020 (figure 13). A similar, albeit less pronounced, trend can be
observed in helpline data relating to the first lockdown in Argentina, with a noticeable increase in call
volumes as restrictions were implemented (figure 14). The patterns observed in Italy and Argentina are
corroborated by recent studies that assessed the impact of COVID-19 confinement measures on violence
against women using call service data.
28
Helpline data from other countries only show minor fluctuations after lockdown measures were introduced,
but they disappeared shortly afterwards. For example, helpline data from Denmark (figure 15) and Mexico
(figure 16) suggest slight decreases in call volumes after confinement measures were introduced, followed
by a return to the pre-lockdown level shortly afterwards. Some helpline data also show changes within the
range of the variability observed in the pre-pandemic period.
29
Overall, the broader impact of COVID-19 restrictions on other types of gender-based violence (beyond
homicidal violence) remains uncertain due to the limited availability of data. Criminal justice data suggest
that there was a predominantly declining trend in the number of incidents of sexual assault and rape
reported to and recorded by authorities shortly after lockdown measures were implemented. This suggests
that during the first wave of confinement measures there was a decrease in the accessibility of women to
institutions for reporting incidents, which often translated into an increase in requests for assistance
directed to helplines for victims of domestic violence.
Weekly number of requests for help from victims of violence received by national helpline,
Italy (February–April 2019 and 2020)
Source: UNODC, elaborated from weekly data from Italy’s national helpline for victims of gender-based violence and
stalking (https://www.1522.eu/).
Note: The first national lockdown was imposed on 9 March 2020 (week 11).
27
In relation to the pre-lockdown period in 2020 and the reference year 2019.
28
Leslie, E. and Wilson, R., “Sheltering in place and domestic violence: evidence from calls for service during COVID-19”, Journal of Public
Economics, vol. 189 (2020); Aguero, J., “COVID-19 and the rise of intimate partner violence”, World Development, vol. 137 (2020).
29
For more information, see UNODC Research brief, “What crime and helpline data say about the impact of the COVID-19 pandemic on
reported violence against women and girls” (2020), p.4.
0
50
100
150
200
250
300
350
400
450
500
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Number of calls per week
Week
COVID-19 confinement measures 2019 2020
Killings of women and girls by their intimate partner or other family members
24
Average daily number of calls received by gender-based violence helpline, Argentina
(March–September 2019 and 2020)
Source: UNODC, elaborated with data from the Ministry of Women, Gender and Diversity, Argentina.
Number of weekly calls received by national helpline, Denmark (10 February–31 May 2020)
Source: UNODC, elaborated from Lev Uden Vold (Live Without Violence), national helpline providing support to victims,
perpetrators, family/friends and professionals, Denmark.
Total number of calls related to incidents of violence against women received by national
helpline, Mexico (January–September 2019 and 2020)
Source: UNODC, elaborated with data from Secretariado ejecutivo del Sistema nacional de seguridad publica, Mexico.
0
50
100
150
200
250
300
March
April
May
June
July
August
September
Number of calls per week
Strict lockdown 20 March - 10 May 2019 2020
0
20
40
60
80
100
120
7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22
Number of calls per week
Week
COVID-19 confinement measures
0
5,000
10,000
15,000
20,000
25,000
30,000
Jan
Feb
Mar
Apr
May
June
July
August
Septemb
er
Number of calls
2019 2020
Killings of women and girls by their intimate partner or other family members
25
Countries can implement different types of policy interventions to prevent gender-related killings and,
more broadly, gender-based violence against women and girls. The evidence presented in this research
brief indicates that this type of violence tends to remain constant over time, even in situations where other
types of lethal violence show signs of decreasing. In line with the call to action by the United Nations
Secretary-General,
30
long-term strategies to prevent gender-related killings by addressing harmful social
norms that normalize violence against women should be combined with measures that can be
implemented in the short term to protect women and girls in situations of risk, especially in situations of
domestic abuse. This can be done by strengthening accessibility to services for female victims of violence
and by intervening in specific situations that can trigger lethal violence among intimate partners and
families.
Four areas of intervention are of key importance in addressing the problem of gender-related killings of
women and girls.
1. Closing data gaps in gender-related killings of women and girls
The collection and dissemination of comprehensive and disaggregated data on gender-related killings
of women and girls is a priority for increasing public awareness and developing evidence-based policies
and programmes.
31
Data on gender-related killings of women and girls are still not available in many
countries and, where they exist, often rely on incomparable national definitions and practices. The
forthcoming statistical framework for measuring gender-related killings, developed jointly by UNODC
and UN Women, can provide a standardized approach and support Member States in going beyond
existing limitations concerning the collection and disaggregation of data on gender-related killings of
women and girls. In addition to female intimate partner/family-related homicide, which is the initial
building block for the measurement of gender-related killings, the statistical framework provides
guidance on identifying and counting other types of homicides related to gender factors. The statistical
framework is aligned to the International Classification of Crime for Statistical Purposes (ICCS),
32
and
can be used by Member States to collect, analyse and disseminate data on different forms of gender-
related killings and on the criminal justice response (arrests, prosecutions, convictions) to such killings.
The statistical framework can also be a tool for strengthening and harmonizing existing efforts to
collect data on gender-related killings by civil society organizations, academia and initiatives such as
Femicide Watch.
33
2. Preventing and addressing gender-based violence by intimate partners or other family members
Considering that most gender-based killings of women and girls are perpetrated by intimate partners
or other family members, specific interventions tackling lethal gender-based violence in the private
sphere are warranted. At least 155 countries have already passed laws on domestic violence,
34
and
most national action plans on gender-based violence or violence against women include specific
actions against domestic and intimate partner violence. A range of interventions can prevent the
escalation of violence against women in domestic contexts, such as:
Protection orders
Protection orders and restraining orders have been found to decrease the likelihood of intimate
partner violence and repeat-victimization after separation.
35
However, after police intervention,
30
Report of the Secretary-General on action against gender-related killing of women and girls (A/70/93). Available at
https://digitallibrary.un.org/record/797541?ln=en.
31
See https://forum.generationequality.org/news/covid-19-exposes-fault-lines-gender-equality-strong-focus-violence-against-women-un-
general.
32
Available at https://www.unodc.org/unodc/en/data-and-analysis/statistics/iccs.html.
33
See A/76/132.
34
World Bank Group, Women, Business and the Law 2020 (Washington D.C., 2020).
35
Holt, V. L. et.al., “Do protection orders affect the likelihood of future partner violence and injury?”, American Journal of
Preventive Medicine, vol. 24, No. 1 (2003), pp. 16-21.
Killings of women and girls by their intimate partner or other family members
26
perpetrators may become more antagonistic and violence can quickly escalate. Further mechanisms
and interventions that ensure protection from violent partners may be necessary for preventing
intimate partner homicide.
36
Simple, quick and accessible protection and restraining or barring orders,
as well as a registration system for such orders, are effective when they form part of a broader package
of measures aimed at preventing lethal risks to women, including confidential 24/7 hotlines and
domestic violence shelters.
37
Domestic violence services
Evidence assessing the impact of domestic violence services (such as helplines and shelters) suggests
that these can reduce victims’ exposure to domestic violence, which is associated with a decline in
intimate partner homicides.
38
It is important to ensure that such support can be sustained over the
long term and, if necessary, combined with other legal measures, such as protection orders, to ensure
that women affected by intimate partner violence have access to places of safety, as well as the option
to remain in their home, based on the principle that the perpetrator and not the survivor should have
to leave the home.
Firearm control
The presence of a firearm in the household has been identified as a risk factor for intimate partner
killings.
39
Limiting perpetrators’ access to firearms, especially in situations when there is previous
domestic abuse and the victim and perpetrator are living in the same house, can prevent intimate
partner violence from culminating in a killing.
40
Awareness-raising campaigns on the risks of exposure
to firearms in domestic disputes should also be considered.
41
Divorce
Divorce legislation should be considered carefully in relation to its possible impact on violence against
women and girls. As divorce can be a period of vulnerability for women,
42
a holistic approach should
be adopted in divorce legislation and coordination among government agencies and courts is needed
for preventing and addressing violence that could lead to gender-related killings of women and girls in
a coherent manner.
43
3. Ensuring effective prosecution and sanctions for perpetrators of gender-related killings
The criminal justice response to gender-related killings should ensure that justice is accessible and
responds to the needs of victims and survivors, and should be commensurate with the gravity of the
offence, holding perpetrators accountable and sending a clear message to society about the gravity of
such killings.
44
In some countries, in particular in Latin America, specific legislation has been passed on
the criminalization of gender-related killing as a measure to strengthen the criminal justice response,
while in other countries, gender-related killings are treated in the same vein as other intentional
36
Dugan, L. et al., “Do domestic violence services save lives?”, National Institute of Justice Journal, Issue 250 (US State Department, 2003).
37
See report of the Secretary-General report on action against gender-related killing of women and girls (A/70/93), para. 41(x); and
Updated Model Strategies and Practical Measures on the Elimination of Violence against Women in the Field of Crime Prevention and
Criminal Justice (A/RES/65/228). Available at: https://undocs.org/a/res/65/228.
38
Dugan, L. et al., “The effects of state and local domestic violence policy on intimate partner homicide”, National Institute of Justice,
(Washington, 2004).
39
See, for example, Campbell J.C. et al., “Risk factors for femicide in abusive relationships: Results from a multisite control study”,
American Journal of Public Health, vol. 93, No. 7 (2003), pp. 1089-1097.
40
See Resolution 65/228 of the United Nations General Assembly “Strengthening crime prevention and criminal justice responses to
violence against women” (A/RES/65/228).
41
See report of the Secretary-General on action against gender-related killing of women and girls (A/70/93), para. 41(v).
42
See UNODC, Strengthening Crime Prevention and Criminal Justice Responses to Violence against Women (Vienna, April 2014).
43
See report of the Secretary-General on action against gender-related killing of women and girls (A/70/93), para. 41(x); and UNODC,
Strengthening Crime Prevention and Criminal Justice Responses to Violence against Women (Vienna, April 2014).
44
UNODC guidance on the criminal justice response to gender-based violence, including gender-based killings, is available at:
https://www.unodc.org/unodc/en/justice-and-prison-reform/cpcj-tools-gender.html.
Killings of women and girls by their intimate partner or other family members
27
homicides.
45
In general, criminal law reform should be accompanied by a review of criminal policies –
including investigatory and prosecutorial policies and practices – to ensure that action is taken with
due diligence and without delay.
46
At the level of criminal justice institutions, this requires the
establishment of specialized units and mechanisms to monitor and sanction criminal justice officials
who do not exercise due diligence in cases of gender-related killing.
47
4. Investing in evidence-based prevention of gender-based violence against women and girls
In the long term, it is necessary to put in place interventions at multiple levels that can reduce the
prevalence of gender-based violence and reduce the number of gender-related killings of women and
girls. Based on a systematic review of different prevention interventions, the United Nations system
has identified seven key strategies to prevent violence against women, summarized in the RESPECT
framework,
48
which can address the root causes of gender-based violence at the country level:
1. Relationship skills strengthened – this strategy refers to interventions aimed at individuals or
groups of women, men or couples for improving skills in interpersonal communication,
conflict management and shared decision-making.
2. Empowerment of women – this strategy refers to both economic and social empowerment
interventions, including inheritance and asset ownership, and microfinance.
3. Services ensured – this strategy refers to a range of services, including police, legal, health,
and social services, provided to survivors.
4. Poverty reduced – this strategy refers to interventions, ranging from cash transfers to labour
force interventions, targeting women and households with the primary aim of alleviating
poverty.
5. Environments made safe – this strategy refers to interventions aimed at creating safe schools,
public spaces and work environments, among others.
6. Child and adolescent abuse prevented – this strategy refers to interventions aimed at
establishing nurturing family relationships, prohibiting corporal punishment and
implementing parenting programmes.
7. Transformed attitudes, beliefs and norms – this strategy refers to interventions that
challenge harmful gender attitudes, beliefs, norms and stereotypes that uphold male
privilege and female subordination, that justify violence against women and that stigmatize
survivors. These may range from public campaigns and group education to community
mobilization efforts.
The seven strategies should not be treated as silos, as successful programmes often have implemented
interventions that fall across more than one of the seven strategies. Only a wide-ranging approach and
long-term commitment can lead to a substantial reduction in the number of gender-related killings.
45
Deus, A., and Gonzales, D., Análisis de Legislación sobre Femicidio/Feminicidio en América Latina y el Caribe e Insumos para una Ley
Modelo, ONU Mujeres (Panamá, 2018).
46
See report of the United Nations Secretary-General on action against gender-related killing of women and girls (A/70/93), paras. 10 and
41 (bb-ee).
47
Ibid., para 41 (ff-jj).
48
WHO, “RESPECT women: Preventing violence against women” (2015). Available at https://www.who.int/publications/i/item/WHO-RHR-
18.19.
Killings of women and girls by their intimate partner or other family members
28
Country-level trends in female intimate partner/family-related homicides in Europe
(2014-2020)
Country-level trends in female intimate partner/family-related homicides in the Americas
(2014-2020)
0
20
40
60
80
100
120
140
160
180
200
2014 2015 2016 2017 2018 2019 2020
Number of female intimate partner/family
-
related homicides relative to baseline (2014)
Country trends indexed Total indexed
0
20
40
60
80
100
120
140
160
180
200
2014 2015 2016 2017 2018 2019 2020
Number of female intimate partner/family
-related
homicides relative to baseline (2014)
Country trends indexed Total indexed
Killings of women and girls by their intimate partner or other family members
29
Rate of female intimate partner/family-related homicide per 100,000 female population in
selected European countries (2014-2020)
Source: UNODC Homicide dataset.
Country‐level trends in female intimate partner/family‐related homicides in selected
countries of the Americas (selected years)
Note: Available data for the United States refer to female intimate partner/family-related homicides and are based on the
FBI’s Supplementary Homicide Reports (available at https://ucr.fbi.gov/crime-in-the-u.s/2019/crime-in-the-u.s.-
2019/tables/expanded-homicide-data-table-10.xls). Available data for Brazil and Mexico refer to “femicide” cases as
defined in respective national legislations. Besides killings of women and girls by intimate partners or other family
members, other types of gender-related killings can be considered as femicides. For Mexico, see Secretariado Ejecutivo del
Sistema Nacional de Seguridad Pública, Gobierno de Mexico (available at
https://www.gob.mx/sesnsp/documentos/lineamientos-para-el-registro-y-clasificacion-de-los-presuntos-delitos-de-
feminicidio?state=published) and for Brazil, see Fórum Brasileiro de Segurança Pública (available at
https://forumseguranca.org.br/wp-content/uploads/2021/07/anuario-2021-completo-v6-bx.pdf). The selection of the
specific criminal offence to conduct investigation and/or prosecution in cases when a woman is intentionally killed is subject
to determination by competent authorities and this can affect comparability and even completeness of relevant data (see
UNODC, Global Study on Homicide 2019). Available data for Honduras and El Salvador refer to female intimate
partner/family-related homicides and are based on the UNODC Homicide dataset.
0.0
0.2
0.4
0.6
0.8
1.0
1.2
2014 2015 2016 2017 2018 2019 2020
Rate of female intimate
partner/femily
-related homicide (per
100,000 female population)
Belarus
Romania
France
Germany
Italy
Spain
United Kingdom
0
20
40
60
80
100
120
2017 2018 2019 2018 2019 2020 2018 2019 2020 2015 2019 2015 2018
United States Brazil Mexico Honduras El Salvador
INDEXED NUMBER OF GENDER
-
RELATED KILLINGS RELATIVE TO
BASELINE
Killings of women and girls by their intimate partner or other family members
30
Number of gender-related violent deaths in India (2016-2020)
Source: National Crime Record Bureau, Ministry of Home Affairs, India.
Trend in the rate of female intimate partner/family-related homicide per 100,000 female
population in Australia and New Zealand (2014-2020)
Source: UNODC Homicide dataset.
Female homicide rates in Kenya, Morocco, South Africa, and Uganda (2010-2020)
Source: UNODC Homicide dataset.Methodological Annex
0
2000
4000
6000
8000
10000
2016 2017 2018 2019 2020
Number of gender
-related violent
deaths
Honour killings
Murders related to witchcraft accusations
Murders related to dowry motives
Dowry deaths (as per Section 304B of the Penal Code)
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
2014 2015 2016 2017 2018 2019 2020
Rate of female intimate
partner/femily
-related homicide
(per 100,000 female population)
Australia New Zealand
0
2
4
6
8
10
12
2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Female homicide rate (per 100,000
female population)
South Africa Uganda Kenya Morocco
Killings of women and girls by their intimate partner or other family members
31
1. Data
1.1. Data sources
The analysis presented in this brief and the estimates produced at global and regional level are based on
data available for 95 countries or territories. For these countries or territories at least one data point on
intentional homicides of female victims perpetrated by their intimate partner or other family members
was available for the period 2010-2020. These data were primarily submitted to UNODC by Member States
through the annual United Nations Surveys on Crime Trends and the Operations of Criminal Justice Systems
(UN-CTS).
49
The UN-CTS collects administrative data on the total number of homicide victims as well as
relevant disaggregations of homicide victims by sex and age, by killing mechanism (firearms, sharp objects
and others) and by perpetrator/context of the crime (family/intimate partner, organized crime, gang,
robbery, other, unknown context). In most cases, the homicide data are sourced from Member States’
criminal justice systems, however, in a few cases the data are sourced from the public health system.
50
Where needed and applicable, UN-CTS homicide data were supplemented with external data sources,
mostly using official sources from Member States or UNODC’s Global initiative to improve knowledge of
the impact of COVID-19 on crime.
51
Population data used to calculate sex-disaggregated homicide rates
(per 100,000 population) were sourced from the United Nations World Population Prospects.
52
1.2. Data validation
All homicide data collected through the UN-CTS as well as external data sources were validated to ensure
that they meet a set of minimal quality criteria, including consistency with the standard definition of
intentional homicide in the International Classification of Crime for Statistical Purposes (ICCS),
53
coherence
of country time trends (to ensure the use of data with consistent counting rules and reporting authorities),
and internal consistency of homicide disaggregations (for example to ensure the male and female
homicides add up to the reported total, or that the reported number of female intimate partner/family-
related homicides does not exceed the reported number of female homicides).
54
1.3. Data coverage
Internationally comparable data on female homicides and female intimate partner/family ‐ related
homicides remain patchy, especially in Africa and Asia, but are available for a large enough number of
countries and territories to enable the production of global and regional estimates. Table 1 provides and
overview of data coverage of these two indicators in the five world regions.
49
Available national data accessible at: https://dataunodc.un.org/content/homicide-country-data
50
See Methodological Annex to The Global Study on Homicide 2019 for more information.
51
More information available at: https://www.unodc.org/unodc/en/data-and-analysis/coronavirus.html
52
Available at: https://population.un.org/wpp/
53
Available at: https://www.unodc.org/unodc/en/data-and-analysis/statistics/iccs.html
54
For a detailed description of the validation procedure applied to the homicide data, see Methodological Annex to The Global Study on
Homicide 2019.
Killings of women and girls by their intimate partner or other family members
32
Table 1: Data coverage for female homicides and female intimate partner/family-related homicides by
region (2010-2020)
Region
Number of
countries/
territories in
region
with available data for at least 1
year in the period 2010-20
with available data for at least 3
years in the period 2010-20
Female
homicide
Female
intimate
partner/family
-related
homicide
55
Female
homicide
Female
intimate
partner/family-
related
homicide
Africa
58
15
6
14
5
Americas
54
49
31
48
26
Asia
51
35
17
31
14
Europe
48
43
37
41
32
Oceania
23
9
4
4
2
World
234
151
95
138
79
Source: UNODC homicide dataset
2. Methods
The methods used for estimating the number of female intimate partner/family-related homicides at the
global and regional level aim to make the best possible use of available data. For each regional aggregate,
the number of female intimate partner/family-related homicides should correspond to the sum of all
national data of such killings in the region, in each year. However, for many countries, data on female
intimate partner/family-related homicides are not available, or data are available only for some years (see
Table 1). As a result, the sample of countries with available data is different for each year. If left
unaddressed, this issue would result in inconsistencies, as regional aggregates would be drawn from a
different set of countries each year.
2.1. Imputation of missing values at the country level
The imputation of missing values at the country level follows a three-step procedure, whereby first, the
values in the total homicide series are imputed if missing, second, the total homicide series is used to
inform the imputation of missing values in the female homicide series,
56
and third, the female homicide
series is then used to inform the imputation of missing values in the female intimate partner/family-related
homicide series.
55
In 8 countries (6 in the Americas and 2 in Asia) data on other types of gender-related killings such as ‘femicides’ or ‘feminicides’ were
used as proxy measures given that reliable data on female intimate partner/family-related homicides are not available. These proxy
measures were primarily used to capture information on country-level trends. For a detailed comparison of data on female intimate
partner/family-related homicides and ‘femicides/feminicides’, see The Global Study on Homicide 2019, Booklet 5, p.25
56
Coverage for female homicide is generally better compared to female intimate partner/family-related homicide, see Table 1.
Killings of women and girls by their intimate partner or other family members
33
1. Imputation of the total homicide series is performed on the country-level rate of total homicides per
100,000 population.
2. Imputation of the female homicide series is performed on the country-level ratio of female homicides
over total homicides (where data on both indicators are available).
3. Imputation of the female intimate partner/family-related homicide series is performed on the
country-level ratio of female intimate partner/family-related homicides over total female homicides
(where data on both indicators are available).
57
For all three series, the following three-step imputation approach is applied:
1. If a country has just one available data point in the respective series,
58
all missing values are set equal
to this single available data point. Given that the estimated series is either a rate (if the imputation
refers to total homicide) or a ratio based on that rate (if the imputation refers to female homicide or
female intimate partner/family-related homicide), this approach accounts for population growth
over time and does not mean that the series is constant in absolute terms.
2. If a country has two to eight available data points in the respective series, the missing values between
two data points are estimated by linear interpolation, and if there are missing values that are
temporally before (or after) the earliest (or latest) available data point, the values at the beginning
(or end) of the series are filled with the earliest (or latest) available data point.
3. If a country has more than eight available data points in the respective time series, the missing values
between two data points are estimated by linear interpolation, and if there are missing values that
are temporally before (or after) the earliest (or latest) available data point, the values at the end of
the time series are imputed using an exponential smoothing approach.
59
2.2. Estimation of regional aggregates
Once the series have been computed at the national level, they are aggregated at the regional level using
the following approach:
• Regional homicide totals are calculated for each year by multiplying the regional homicide rate per
100,000 population with the total population of the respective region (divided by 100,000).
60
• Regional female homicide totals are calculated for each year by multiplying the regional ratios of
female homicides over total homicides with the total homicides of the respective region.
• Regional totals of female intimate partner/family-related homicide are calculated for each year by
applying the regional ratios of female intimate partner/family-related homicides over total female
homicides to the total female homicide series of the respective region.
Finally, regional estimates are aggregated to compute the global number of female intimate
partner/family-related homicides.
2.3. Computation of uncertainty intervals
As explained above, global and regional estimates of female intimate partner/family-related homicides are
produced based on available national data and through a statistical model imputing missing values at the
country level. To account for estimation error due to the imputation of missing values at the country level,
57
Validation checks ensure that this ratio is lower or equal to one.
58
The three-step imputation approach is applied to the period 2000-2020.
59
For more information, see https://afit-r.github.io/ts_exp_smoothing
60
In order to better reflect the diversity of homicide levels across African countries a different approach was used to calculate the regional
homicide estimate for Africa, which rests on the classification of countries into three groups according to levels of insecurity. For more
information, see Methodological Annex to The Global Study on Homicide 2019, p.viii.
Killings of women and girls by their intimate partner or other family members
34
the global and regional estimates are accompanied by intervals of uncertainty. These bands are intended
to represent how the uncertainty due to imputation varies across time and regions. It is important to note
that these bands do not represent confidence intervals and should not be interpreted as such. The
procedure to compute the uncertainty intervals closely follows the approach used in the Global Study on
Homicide 2019.
61
For countries and territories with at least one year of data on female intimate partner/family-related
homicide (in the period 2010-2020), a penalty of plus and minus 2.6 per cent was added
62
to the value of
each estimate for each year of distance between the estimate and the closest observed female intimate
partner/family-related homicide value. For countries and territories without any reported data on female
intimate partner/family-related homicide, a maximum “penalty” is applied in the sense that the distance
to the closest observed values is assumed to be 10 years. Ranges around global and regional trends were
obtained by adding up – for each year – the compounded ranges of all countries that are a part of each
regional aggregate. This approach to calculating uncertainty intervals means that countries with fewer
years of reported data have wider ranges around the national trend and contribute more to the ranges
around the trends of their respective regions. In addition, regions with a greater number of countries
without any reported data have even wider ranges.
61
See Methodological Annex to The Global Study on Homicide 2019, p.ix
62
The 2.6 per cent represent the global average change in the ratio of female intimate partner/family-related homicides over total female
homicides. For the two regions with sufficient data coverage, Europe and the Americas, the regional average in the ratio of female
intimate partner/family-related homicides over total female homicides is used instead (Europe: 2.6 per cent; Americas: 3.1 per cent).
