Available via license: CC BY 4.0
Content may be subject to copyright.
Contents lists available at ScienceDirect
Journal of Forensic and Legal Medicine
journal homepage: www.elsevier.com/locate/yjflm
Three groups of suspects in police reported rape cases: First-time suspects,
recidivists and unidentified suspects. A comparative study
Bjarte Frode Vik
a,b,∗
, Kirsten Rasmussen
c,d
, Berit Schei
e,f
, Cecilie Therese Hagemann
a,f
a
Norwegian University of Science and Technology (NTNU), Department of Clinical and Molecular Medicine, P.B. 8905, N-7491, Trondheim, Norway
b
St. Olavs Hospital, Department of Neuropsychiatry, P.B. 3250 Sluppen, N-7006, Trondheim, Norway
c
Norwegian University of Science and Technology (NTNU), Department of Psychology, 7491, Trondheim, Norway
d
St. Olavs Hospital, Forensic Research Unit, Brøset, P.B. 1803 Lade, 7440, Trondheim, Norway
e
Department of Public Health and Nursing, Norwegian University of Science and Technology, P.B. 8905, N-7491 Trondheim, Norway
f
Department of Obstetrics and Gynecology St. Olavs Hospital, P.B. 3250 Sluppen, N-7006, Trondheim, Norway
ABSTRACT
Background: Previous studies show that reported suspects in adult rape cases often have a criminal record, and that many are rape recidivists. Annual numbers of
police reported rapes have dramatically increased but the proportion of rapes being prosecuted and numbers of convictions are low. To increase knowledge about the
suspects in cases of police reported rapes; whether they have committed the crime before or not may inform preventive measures.
Aims: To compare suspect, victim, and assault related characteristics among different groups of police-reported rape suspects (first-time suspects, recidivist suspects
and unidentified suspects).
Methods: Retrospective, descriptive study of suspects in cases of rape or attempted rape reported by women ≥16 years of age in the Sør-Trøndelag police district,
Norway, from 2003 to 2010.
Results: Among the 356 suspects included, 207 (58%) were first-time suspects, 75 (21%) were recidivists and 74 (21%) were unidentified. Being a first-time suspect
was significantly associated with victim being <18 years, recidivist suspect was significantly associated with victim being a partner, both suspect- and victim
unemployment, and suspect reporting intake of other drugs than alcohol. When suspects were unidentified, victims were more likely to have consumed alcohol prior
to assault, and reporting the suspect being of non-Western origin. Also, the reporting of a public venue was more frequent when unidentified suspect.
Conclusions: The study shows different patterns in groups of suspects as to victim and assault characteristics. Detection and description of such differences can
provide valuable information for future prevention programs, police investigation methods and health care guidelines.
1. Introduction
There has been a steady increase in police-reported rapes in
Norway, from 400 per year in the 1990s to almost 1600 per year in
2016.
1,2
Despite increasing rates of police-reported rapes, the percen-
tage of cases proceeding to prosecution is low and even decreasing in
Norway, in line with findings from other countries.
3,4
During the last 30 years a major focus of research and policy im-
plications has been on the sexual recidivists and the prevention of
sexual recidivism.
5,6
This is a result of the perception that sexual of-
fending is a life-course persistent inclination, and it has led politicians
to seek predominantly punitive solutions to a problem as complex as
sexual violence. New research underlines the importance of under-
standing the origin and the development of sexual offending over time
and the factors responsible for it, to better understand and prevent
sexual recidivism in the future.
7
It has also been emphasized that the
attention regarding research on prevention of sexual assault and of-
fender treatment should be drawn towards youths who commit sexual
crimes, as all data point to the origin of sexual offending in the early
adolescent years.
8
Traditional theories regarding sexual assault postu-
late that sexual offenders specialize in types of victims and/or offences
and are categorized in certain typologies.
5,9,10
Research has shown that
rapists often have many previous convictions for a violent crime, and
that they resemble violent offenders or criminals in general. In contrast,
those convicted for sexual crimes against children tend to engage in
sexual offending exclusively.
5,11
There are few characteristics, traits or
patterns of human behavior which can be used to generalize hallmarks
in a population of sexual assailants.
12
On the other hand, according to a
report from the Oslo Police District (OPD), the population of sexual
offenders are different from others in several ways.
13
They are both
more victimized and victimizing, they have more mental health pro-
blems and have far more often a criminal record in the police files than
the general male population. Two thirds of the persons reported to the
OPD for rape in the year 2010 had formerly registered criminal activity
in police files. In the period 2000–2010 the OPD registered an increase
in sexual assaults committed by men of non-European descent. The
https://doi.org/10.1016/j.jflm.2019.02.004
Received 6 June 2018; Received in revised form 30 October 2018; Accepted 10 February 2019
∗
Corresponding author. St. Olavs University Hospital, Dept. of Neuropsychiatry, Trondheim University Hospital, P.B. 3250 Sluppen, Trondheim, N-7006, Norway.
E-mail address: bjarte.vik@ntnu.no (B.F. Vik).
Journal of Forensic and Legal Medicine 71 (2020) 101771
Available online 15 February 2019
1752-928X/ © 2019 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license
(http://creativecommons.org/licenses/by/4.0/).
T
authors emphasized that victims' threshold for reporting to the police
was probably lower when the assailant's origin was non-European.
13
Little is known about how the police prioritize in investigations of
sexual assault cases. There is some evidence showing that preconceived
attitudes in society towards sexual assault victims influence on how the
police handle sex offences.
3
Theories on rape myths were introduced in
the 1970s to explain false beliefs about how and why women are
sexually assaulted.
14
Research has documented that law enforcement
personnel endorse rape myths.
15–17
Recent research has shown that
rape myths are documented in official rape case records and suggests
that this may influence investigative responses and perhaps predict case
progression in a negative way.
18
Preconceived attitudes are also de-
scribed as existing towards sexual assault suspects. A British study
showed that suspects who had a criminal record, especially as sex of-
fenders, and those of non-white skin color, had an increased risk of
getting convicted.
19
“The credible criminal” was the term used to de-
scribe these offenders and the findings were later supported by a Danish
study.
3
However, research in this area is limited. The first aim of this
study was to describe and to compare the following three groups of
suspects of rape or attempted rape in the Sør-Trøndelag Police District
(STPD): 1) Suspects who were police-reported for the first time (re-
ferred to as first-time suspects). 2) Suspects who had one or more
former episodes registered in the STPD files as a suspect of rape or other
violent crime (referred to as recidivists), and 3) Unidentified suspects.
The descriptions and comparisons of the suspect groups were based on
victim-, assailant and assault characteristics, in police reported rapes
and attempted rapes. A second aim was to describe differences in police
investigations and prosecution rates of rape cases depending on the
group of suspect/assailant involved.
Fig. 1. Flow chart for included and excluded police-reported cases of rape and attempted rape in Sør- Trøndelag police district during the period 2003–2010. Police
data are merged with data from the Trondheim SAC in corresponding cases of sexual assault.
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
2
2. Material and methods
2.1. Study design and settings
We conducted a retrospective descriptive study of police-reported
rapes and attempted rapes of female victims ≥16 years of age in the
Sør-Trøndelag Police District (STPD) in Norway between July 1, 2003
and December 31, 2010. The population of the area is approximately
280 000, including the city of Trondheim with about 160 000 in-
habitants.
20
The only medical sexual assault center (SAC) in the district
is located at St. Olavs Hospital in Trondheim. Data from police files
were merged with corresponding medical information from the SAC
when available. The SAC's service is described in detail elsewhere.
21
The Institute of Forensic Medicine, Oslo University, carries out all for-
ensic analyses of collected biological samples from sexual assault vic-
tims in Norway. The results from these analyses are kept in police files,
and are usually unavailable for the SAC personnel.
22
2.2. Participants
All police-reported cases of rape and attempted rape of women ≥16
years of age, were identified. Cases were selected based on the former
Norwegian Penal Code.
23
According to this law, Chapter 19, section
192, rape was defined as in the following abbreviated version: Pene-
tration of penis/finger/foreign object in vagina/anus, penis in mouth,
masturbation, and coercion by means of violence, threats, or during
impaired consciousness.
23,24
Four paragraphs under section 192 cov-
ered various distinctions of rape: Among the 320 suspects of rape in-
cluded in this study, 306 were registered according to section 192,
paragraphs 1 and 2 (rape), 9 suspects according to paragraph 3 (ag-
gravated rape) and 5 suspects according to paragraph 4 (grossly neg-
ligent rape). Attempted rape, which was registered in 36 suspects in the
study, is also punishable, but is covered by another paragraph in the
Penal Code.
A total of 475 cases were police-reported during the period. Cases
were excluded according to Fig. 1. Male victims (n= 18), those < 16
years of age (n= 49), unidentified victims (n= 3) and duplicate re-
gistrations (n= 21) were excluded. Additionally, some cases were ex-
cluded because of missing information regarding whether suspects were
previously registered as suspects in police files or not (n= 28), leaving
356 cases eligible for the study. Among the 356 cases, we had corre-
sponding medical information from the SAC in 212 cases (Tables 1 and
3and Fig. 1).
2.3. Data collection and variables
Characteristics of suspects and victims were retrieved from police
files, but in cases where SAC information was available and information
in the police files was missing, victims' medical records were the source
of information. We collected data on the following variables: Suspects'
and victims' sociodemographics like age, origin (classified as Western if
stated as Western Europe, North America or Oceania (Australia or New-
Zealand), else classified as Non-Western), occupational status, and
education. Age of unidentified suspects was based on victims' self-re-
porting. Furthermore, suspect-victim relationship was categorized into
partner (current or previous partner/husband/boyfriend), family
member or friend, casual contact (suspect known < 24 h) or stranger
(suspect not previously known). Type of sexual acts involved anal,
vaginal or oral penetration, or recorded as “no recollection/missing”.
Use of physical violence was graded as severe (presence of weapon/
attempted strangulation/fracture or internal injuries), light/moderate
(holding/punch/kick) or none/verbal threats. Location of assault/
venue was categorized into private (woman's, assailant's or other per-
son's residence) and public (any public indoor or outdoor location or a
vehicle). Victim alcohol intake in relation to the assault was classified
by three categories; “no intake”, “intake of <5 units of alcohol” and
“intake of ≥5 units of alcohol/heavily intoxicated”. Suspect alcohol
intake was dichotimized to “yes” or “no”. In cases with more than one
assailant, information regarding the presumably most active one was
used. In cases with unidentified suspect, information given by the
victim or other witnesses was used.
Table 1
Suspect, victim, assault and investigative characteristics and by former suspects
of sexual assault or other violent crime, 2003–2010.
Variable Total
N = 356
n(%)
First-
time
suspects
N = 207
n (%)
Recidivists
N = 75
n (%)
Unidentified
suspects
N = 74
n (%)
p-value
Suspect characteristics
Origin
Western 235 (66) 150 (73) 57 (76) 28 (38) 0.004
a
Non-western 98 (28) 55 (27) 17 (23) 26 (35)
Missing 23 (7) 2 (1) 1 (1) 20 (27)
Victim characteristics
Age
16–17 years 72 (20) 53 (26) 8 (11) 11 (15) 0.02
b
18–24 years 160 (45) 80 (39) 42 (56) 38 (52)
≥25 years 123 (35) 73 (35) 25 (33) 25 (34)
Missing 1 (0) 1 (1) 0 (0) 0 (0)
Occupation
Employed/
student
219 (62) 132 (64) 41 (55) 46 (62) 0.1
a
Unemployed 64 (18) 33 (16) 20 (27) 11 (15)
Missing 73 (21) 42 (20) 14 (19) 17 (23)
Victim alcohol intake
No intake 103 (29) 65 (31) 25 (33) 13 (18) 0.007
b
<5 units 67 (19) 36 (17) 8 (11) 23 (31)
>5 units 157 (44) 89 (43) 38 (51) 30 (41)
Missing 29 (8) 17 (8) 4 (5) 8 (11)
Assault characteristics
Victim-suspect relationship
Partner 57 (16) 38 (18) 19 (25) 0 (0) < 0.001
c
Friend/family 134 (38) 97 (47) 29 (39) 8 (11)
Casual contact 107 (30) 66 (32) 19 (25) 22 (30)
Stranger 55 (15) 6 (3) 8 (11) 41 (55)
Missing 3 (1) 0 (0) 0 (0) 3 (4)
Penetration
No penetration 70 (20) 31 (15) 17 (23) 22 (30) 0.006
a
Penetration by
penis or
foreign
object
249 (70) 155 (75) 55 (73) 39 (53)
Missing 37 (10) 21 (10) 3 (4) 13 (18)
Physical violence
None/verbal 45 (13) 32 (16) 10 (13) 3 (4) 0.01
b
Light/moderate 221 (62) 124 (60) 50 (67) 47 (64)
Severe 44 (12) 17 (8) 12 (16) 15 (20)
Missing 46 (13) 34 (16) 3 (4) 9 (12)
Place of assault
Private 235 (66) 157 (76) 56 (75) 22 (30) < 0.001
a
Public 113 (32) 46 (22) 19 (25) 48 (65)
Missing 8 (2) 4 (2) 0 (0) 4 (5)
More than 1 suspect
1 suspect 316 (89) 185 (89) 70 (93) 61 (82) 0.24
a
More than 1
suspect
38 (11) 22 (11) 5 (7) 11 (15)
Missing 2 (1) 0 (0) 0 (0) 2 (3)
Investigational data
Other witnesses interrogated
No 101 (28) 53 (26) 19 (25) 29 (39) 0.08
a
Yes 251 (71) 150 (73) 56 (75) 45 (61)
Missing 4 (1) 4 (2) 0 (0) 0 (0)
Police investigation of crime scene
No 204 (57) 120 (58) 39 (52) 45 (61) 0.5
a
Yes 151 (42) 87 (42) 36 (48) 28 (38)
Missing 1 (0) 0 (0) 0 (0) 1 (1)
a
Chi square, df = 2.
b
Chi square, df = 4.
c
Chi square, df = 6.
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
3
For victims who had been at the SAC, medical records provided data
regarding extragenital and anogenital injuries (the latter included tears,
abrasions and bruises (ecchymoses/petecchiae), and reported redness
or swelling was excluded),
25
collection of biological trace evidence and
results of victims’ toxicological analyses.
Police files provided data on investigational issues like interroga-
tions performed and results from these, collection of medical informa-
tion, including forensic and suspects’ toxicological analyses, results
from these, and legal outcome. Data from police- and medical records
were merged and registered through a web-based data collection
system developed and administered by the Unit of Applied Clinical
Research at the Norwegian University of Science and Technology. In
case of discrepancy between the two data sources information from the
police files were preferred.
The study was approved by the Regional Committee for Medical and
Health Research Ethics, the Norwegian Director General of Public
Prosecutions and the Advisory Board on Secrecy and Research. The
merging of data was also approved by the Norwegian Data Inspectorate.
2.4. Statistical analysis
We compared suspect-, victim- and assault characteristics, as well as
police investigations between the three groups of suspects described
above. For the comparisons, Pearson's chi-square and Exact uncondi-
tional tests were used as appropriate for the categorical variables, and
ANOVA and student's t-test for the continuous variables (age).
Statistical significance was assumed when p< 0.05. Missing data were
calculated but excluded when statistical tests were performed. Data
analyses were performed using IBM SPSS Statistics for Windows, ver-
sion 22.0.
3. Results
3.1. Suspect characteristics (Tables 1 and 2)
Among the 356 cases included in the study, 207 were first-time
suspects (58%), 75 were recidivists (21%) and 74 were unidentified
(21%). The mean age of identified suspects (first-time suspects and
recidivists) was 30.4 years (SD =10.9), ranging from 16 to 84 years.
The mean age of unidentified suspects, based on victims' self-reporting,
was 29.5 years, ranging from 18 to 58 years. (SD = 8.2).
Among unidentified suspects 35% were reported as non-Western,
whereas the corresponding percentages were 27% and 23% among
first-time suspects and recidivists (X
2
= 11.3, df = 2, p = 0.004)
(Table 1).
Among the identified suspects shown in Table 2 the unemployment
rate among first-time suspects was 10%, versus 28% among recidivists
(X
2
= 11.1, df = 1, p = 0.004). One third of the identified suspects had
less than 13 years of education, regardless of suspects being first-time
suspects or recidivists. However, information was often missing re-
garding suspect education in the police records. Suspect alcohol con-
sumption prior to the assault was reported by 60% of the first-time
Table 2
Suspect and investigative characteristics, by identified former suspect of sexual
assault or other violent crime, 2003–2010.
Total
N = 282 n
(%)
First-time
suspects
N = 207 n (%)
Recidivists
N = 75 n (%)
p-value
Suspect characteristics
Age
≤24 years 107 (38) 78 (38) 29 (39) 0.1
a
25–34 years 91 (32) 66 (32) 25 (33)
≥35 years 84 (30) 63 (30) 21 (28)
Occupation
Employed/
student
181 (64) 136 (66) 45 (60) 0.004
b
Unemployed 42 (15) 21 (10) 21 (28)
missing 59 (21) 50 (24) 9 (12)
Education
<13 years 88 (31) 62 (30) 26 (35) 1.0
b
University/
college
27 (10) 19 (9) 8 (11)
Missing 167 (59) 126 (61) 41 (55)
Self-reported alcohol intake
No 81 (29) 53 (26) 28 (37) 0.09
b
Yes 164 (58) 124 (60) 40 (53)
Missing 37 (13) 30 (15) 7 (9)
Self-reported intake of other drugs than alcohol
No 189 (67) 147 (71) 42 (56) 0.015
b
Yes 24 (9) 13 (6) 11 (15)
Missing 69 (25) 47 (23) 22 (29)
Investigational and legal data
Suspect interrogated
No 25 (9) 23 (11) 2 (3) 0.03
b
Yes 248 (88) 176 (85) 72 (96)
Missing 9 (3) 8 (4) 1 (1)
Suspect admits sexual contact
No 76 (27) 60 (29) 16 (21) 0.09
b
Yes 166 (59) 113 (54) 53 (71)
Missing 40 (14) 34 (16) 6 (8)
Suspect DNA-profile taken
No 150 (53) 111 (54) 39 (52) 0.27
b
yes 105 (37) 71 (34) 34 (45)
missing 27 (10) 25 (12) 2 (3)
Suspect arrested
No 188 (67) 137 (66) 51 (68) 1.0
b
Yes 78 (28) 57 (28) 21 (28)
Missing 16 (6) 13 (6) 3 (4)
Prosecution
Yes 32 (11) 19 (9) 13 (17) 0.06
c
Dismissal 239 (85) 179 (86) 60 (80)
Missing 11 (4) 9 (4) 2 (3)
a
Chi square, df = 2.
b
Chi square, df = 1.
c
Exact unconditional test.
Table 3
SAC medico-legal information, by former suspects of sexual assault or other
violent crime. Data only from the Trondheim SAC 2003–2010.
Variable Total
N = 212
n (%)
First-time
suspects
N = 118 n
(%)
Recidivists
N = 48 n (%)
Unidentified
suspects
N = 46 n (%)
p-value
Victim medico-legal findings
Bodily injury (extragenital)
No 71 (34) 35 (30) 21 (44) 15 (33) 0.2
a
Yes 126 (59) 73 (62) 24 (50) 29 (63)
Mising 15 (7) 10 (9) 3 (6) 2 (4)
Anogenital injury
No 140 (66) 80 (68) 34 (71) 26 (57) 0.5
a
Yes 53 (25) 28 (24) 11 (23) 14 (30)
Missing 19 (9) 10 (9) 3 (6) 6 (13)
Tox. samples from victim
No 107 (51) 50 (42) 31 (65) 26 (57) 0.02
a
Yes 105 (50) 68 (58) 17 (35) 20 (44)
Tox. results victim
No tox. agents 31 (15) 22 (19) 7 (15) 2 (4) 0.08
a
≥1 tox. agent 74 (35) 46 (39) 10 (21) 18 (39)
Time from assault to med. exam
<24 h 154 (73) 83 (70) 33 (69) 38 (83) 0.2
a
≥24 h 57 (27) 34 (29) 15 (31) 8 (17)
Missing 1 (1) 1 (1) 0 (0) 0 (0)
Investigational and legal data
Police requested forensic medical record from SAC
No 39 (18) 20 (17) 11 (23) 8 (17) 0.9
a
Yes 103 (49) 57 (48) 24 (50) 22 (48)
Missing 70 (33) 41 (35) 13 (27) 16 (35)
Analysis of swabs and/or clothes collected from victim
No/missing 115 (54) 70 (59) 31 (65) 14 (30) 0.001
a
Yes 97 (46) 48 (41) 17 (35) 32 (70)
a
Chi square, df = 2.
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
4
suspects and 53% of the recidivists (X
2
= 2.8, df = 1, p = 0.09). Use of
other drugs than alcohol in relation to the assault was reported by 6% of
the first-time suspects and 15% of the recidivists (X
2
= 6.4, df = 1,
p = 0.015) (Table 2).
3.2. Victim characteristics (Table 1)
Table 1 also shows victim characteristics by suspect category. Mean
age of the victims was 25.3 years (SD = 9.6), ranging from 16 to 72
years. There was an association between first-time suspects and victims
in the youngest age category (16–17 years), whereas victims assaulted
by recidivists were somewhat older (X
2
= 12.0, df = 4, p = 0.02). The
victim was of non-Western origin in only 3% of the cases. Most of the
victims were employed and/or students (62%). Victim unemployment
was more prevalent in cases with a recidivist suspect than in cases with
a first-time suspect (27% versus 16%) (X
2
= 4.6, df = 2, p = 0.1). In
63% of the cases the victims had consumed alcohol prior to the assault.
Victim alcohol consumption was associated with unidentified suspect
(X
2
= 5.4, df = 4, p = 0.07). Although the percentages of victims who
reported having consumed alcohol prior to the assault were about the
same in the two groups of identified suspects (first-time suspects and
recidivist suspects), there was an association between victim being
highly intoxicated by alcohol (consumed > 5 units) and recidivist sus-
pect.
3.3. Assault characteristics (Table 1)
The different assault characteristics by suspect category are shown
in Table 1. The victim knew the suspect in almost two thirds of the cases
with identified suspect, regardless of whether the suspect was a first-
time or a recidivist suspect. Being a first-time suspect was associated
with cases where the victim was a casual contact (known <24 h) while
recidivist suspect was associated with partner rape. There was also a
higher occurrence of stranger rapes among the group of recidivist sus-
pects than among first-time suspects (X
2
= 138, df = 6, p < 0.001).
Among the 74 unidentified suspects, 22 were classified as casual con-
tact while 41 were strangers.
Among the identified suspects 210 (75%) were accused of a pene-
trative assault, whereas penetration was reported in only 39 (53%) of
the unidentified suspects, (X
2
= 10.2, df = 2, p = 0.006). The victims
of recidivists more often reported to be exposed to physical violence
than victims of first-time suspects (83% versus 68%) (X
2
= 12.6,
df = 4, p = 0.01). Unidentified suspect was associated with a public
venue, while three of four assaults committed by identified suspects
occurred in a private place (X
2
= 52, df = 2, p < 0.001).
3.4. Victim injury and laboratory findings (SAC information, Table 3)
SAC information by suspect category is shown in Table 3. Among
those victims who had been examined at the SAC, extragenital injury
was registered in 59%, while anogenital injury was disclosed in 25% of
the victims. There was no significant association between injury and
suspect category. Half of the victims who attended the SAC had a tox-
icological blood sample collected; in 58% of victims of first-time sus-
pects, in 35% of victims of recidivists and in 44% of victims of uni-
dentified suspect (X
2
= 7.6, df = 2, p = 0.02). Samples disclosed ≥1
toxicological agent in 39% of victims of first-time suspects, in 21% of
victims of recidivists and in 39% of victims of unidentified suspects
(X
2
= 5.0, df = 2, p = 0.08). Around 70% of the victims at the SAC
were examined within 24 h after the assault when the suspect was
identified, this in contrast to when the suspect was unidentified where
83% of the victims came within 24 h (X
2
= 2.8, df = 2, p = 0.24).
3.5. Investigational and legal data (Tables 1–3)
The police requested a forensic medical record from the SAC in half
of the cases in which victims had been medically examined, and there
were no differences between the groups of suspects regarding that
variable (Table 3). Analysis of swabs and/or clothes collected from
victims was conducted in 46% of the cases; in 70% of cases with uni-
dentified suspect, whereas only in 41% and 35% of first-time suspect
cases and recidivist cases, respectively (X2 = 13.8, df = 2, p = 0.001)
(Table 3).
Nearly all victims were interrogated by the police, but in cases
where suspect was identified, the police interrogated other witnesses
than the victim more often than in cases where suspect was unidentified
(75% vs. 61%) (X
2
= 5.1, df = 2, p = 0.08) (Table 1). Interrogations of
suspects were conducted in a significantly higher proportion of re-
cidivists than among first-time suspects (96 vs. 85%, X
2
= 5.1, df = 1,
p = 0.03) (Table 2). Among recidivists, 71% admitted sexual contact
with the victim, whereas 54% of the first-time suspects admitted sexual
contact (X
2
= 3.0, df = 1, p = 0.09) (Table 2). Only 2% of the suspects
admitted rape or culpability and there were no differences in the groups
of suspects regarding these two variables.
A DNA profile of the suspect was secured during investigations in
somewhat more of the recidivist cases than in the first-time suspect
cases (X
2
= 1.2, df = 1, p = 0.27) (Table 2). In 42% of the cases the
police investigated the venue (Table 1).
During investigations, 28% of the identified suspects were arrested
regardless of group of suspects involved (Table 2). Prosecution hap-
pened in 32 cases and was associated with recidivist suspects, of which
17% of the cases were prosecuted in court, whereas only 9% of the first-
time suspect cases were prosecuted (Exact unconditional test, p = 0.06)
(Table 2). A medical doctor from the SAC was summoned as an expert
witness in only five of the 32 cases.
4. Discussion
In this study of police reported rape suspects, 207 of 356 were first-
time suspects (58%), 75 were recidivists (21%) of prior sexual or vio-
lent crime, and 74 were unidentified suspects (21%). We found that the
mean age of identified suspects was 30.4 years. The mean age of uni-
dentified suspects, which was based on self-reporting from victims, was
29.5 years. This corresponds with findings reported by the OPD, that
the mean age of offenders has been approximately 30 years throughout
the decade 2000–2010.
13
The mean age of victims was 25.3 years. In a
Danish study which also included merged data from police files and a
SAC, victim mean age was 26 years.
4
We found that first-time suspects reported higher alcohol intake
prior to the assault compared to recidivists. Their corresponding victims
were also often <18 years and an acquaintance. Being a first-time
suspect was associated with victims reporting less physical violence
than in cases of recidivists and unidentified suspects. These findings
disclose a pattern of assault characteristics that are often seen in what
the Norwegian police categorizes as party-related rapes. According to the
OPD this category of rapes typically occurs when young people parti-
cipate in parties or social events as part of the weekend night life, where
alcohol, and often large amounts of it, is involved.
13
Our finding of
association between first-time suspect and high alcohol intake in sus-
pect is consistent with the OPD report, which describes that the group
of offenders involved in party-related rapes, often has no prior criminal
record.
13
In Norway, episodic heavy drinking is common and to a
certain degree accepted in social settings, regardless of gender. Out of
this we might consider that a possible consequence of this “Nordic
pattern of drinking” could be the effect of disclaiming from responsi-
bility of what happens under the influence of alcohol. Attitudes like
these may contribute to trivializing the seriousness of sexual assaults
happening in such settings, and maybe even partly explain why the
police seem to put less investigational effort in first-time suspect cases
compared with recidivist cases. Our findings of lower prosecution rates
in this group compared to the recidivists is supported by a criminolo-
gist, describing that suspects who have a criminal record tend to get a
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
5
higher priority in criminal investigations than those who do not.
26
Hypothetically an explanation could be that police officers may ex-
perience more stress or discomfort when initiating full investigations of
suspects who do not have a criminal record, compared with in-
vestigating men who are already registered as criminals. As mentioned
earlier, there is research describing preconceived attitudes in society
towards both victims and assailants of sexual offences, which can
contribute to explaining how law enforcement prioritize when in-
vestigating sex crimes.
3,19
Literature is, however, sparse on this topic
and more research is needed.
We found that recidivist suspect was associated with suspect and
victim unemployment, lower suspect education, and suspect reporting
intake of other drugs than alcohol. Our findings indicate that both re-
cidivists and their victims seem to represent relatively vulnerable po-
pulations in the society. Compared to first-time suspects, the recidivists
more often were accused of partner rape and the use of physical vio-
lence. Many studies describe stranger rapes as the more violent cate-
gory when compared to rapes committed by known assailants, but one
study from a Swedish SAC showed that women assaulted by their in-
timate partners were even more exposed to physical violence than
women assaulted by strangers and acquaintances.
27
One interesting
finding in our study was that while 25% of the recidivist suspects were
categorized as partners, 11% were reported as strangers, both cate-
gories far more common than among the first-time suspects. This is
supported by a research group from Finland which showed that a
considerable proportion of men who commit attack rapes are, or have
been, in a steady intimate relationship for a long time.
28
Based on this a
researcher has questioned whether it is likely that some cases of
partner- and stranger rapes might have been committed by the same
perpetrator.
10
Based on our findings, we can even ask whether it is
possible that some men who are in intimate relationships, occasionally
attack not only their partner/spouse, but also other random women.
Previous research has shown that crossover-offending is common
among sex offenders, meaning that many admit to multiple victims and
offences atypical of traditional criminal classification. This verifies that
the theories of sex offender typologies are complex and have limita-
tions.
5
Our results show that the police have been somewhat more thor-
ough in their investigational work regarding recidivist suspects than in
cases of first-time suspects. Interrogations of suspects were more often
done, a DNA-profile was more often secured, and the venue was more
often investigated in the recidivist cases. Almost none of the suspects,
regardless of group, admitted rape, but recidivist suspects more often
admitted sexual contact with the victim, and recidivist cases also more
often ended with prosecution. In most societies a small group of people
commit a large proportion of the crimes, the so-called “acquaintances of
the police”.
26
A Norwegian researcher has analyzed how law enforce-
ment systematically follows and goes after citizens who habitually
violate the law. She describes and justifies the police's close follow-up
and sheds light on how priorities are made in crime investigations in
general. In accordance with our findings, she explains why police in-
vestigations seem to be of higher quality in the cases of recidivist rape
suspects compared with the two other groups of suspects. This phe-
nomenon could also justify the higher prosecution rate in recidivist rape
cases in our study. It is important to communicate this finding to the
public for the purpose of helping victims of possible recidivist sexual
assailants realize that police reporting has a relatively larger potential
of bringing their assailants to court. More research is needed to confirm
our findings regarding various degrees of police investigational efforts
in separate groups of sexual assailants. As expected, most of the sus-
pects registered as unidentified were either a stranger to the victim or a
casual contact (known < 24 h). Cases with unidentified suspect was
associated with non-Western suspect, high victim alcohol consumption
prior to assault, a public venue. This category of rapes draws a picture
of relatively dramatic and frightening assaults. Previous research has
concluded that sexual assault against young women who are too
intoxicated to resist, due to heavy episodic drinking, is a prevalent
problem in Norway.
29
Although most victims in the total sample re-
ported alcohol intake prior to the assault (63%), there was a significant
association between unidentified suspect and victim alcohol consump-
tion (71%). It seems that unidentified assailants are exploiting women
who are incapacitated by alcohol. A study from the U.S. about uni-
dentified sexual assailants supports our findings, describing that ex-
ploitation of the victim's intoxication is a commonly used means of
coercion.
30
The association between unidentified suspect and the use of
physical violence was extracted not only from victim interrogations, but
also supported from SAC medical data, disclosing high percentages of
both extragenital and anogenital injury.
Victims of unidentified suspects also tended to seek help at the SAC
relatively shortly after the incidence, which may be explained through
the physically and psychologically traumatizing nature of these as-
saults. It can also partly be explained through a British study showing
that the threshold for seeking help and police-reporting a rape is lower
when the suspect is a stranger to the victim.
31
Analysis of swabs and/or
clothes collected from the victim was done significantly more often in
cases where suspect was unidentified than in cases of identified sus-
pects, which also presumably has to do with the higher percentage of
these victims being medically examined within 24 h after the rape in-
cident. Despite these analyses we ended up with a rather large group of
unidentified suspects. However, the collected and stored DNA from
unidentified males could be valuable evidence in future investigations.
The unidentified suspects were also more often suspected of attempt of
rape rather than rape, and penetration was less often reported, which is
consistent with previous research concluding that stranger rapists
seldom manage to complete the rape with penetration and ejacula-
tion.
12
We found an association between unidentified suspect and non-
Western suspect. The OPD has showed an increasing occurrence of
suspects of non-Norwegian/non-European origin in the period
2000–2010.
13
Correspondingly, men of non-European origin committed
most of the so-called attack rapes, in which the assailant was uni-
dentified. The numbers were, however, too small to constitute statisti-
cally significant results on that matter.
13
There is little research on how
or why a foreign descent or a background from a different culture is
connected to rape, and explanations to this could to our knowledge only
be based on speculation.
12
We can, however, assume that persons who
originate from foreign countries may have a more difficult social si-
tuation in Norway than ethnic Norwegians. Rapes reported to the police
probably only account for 10% of the total amount of sexual coercions
happening in the society.
32
The OPD claims that dark-skinned Norwe-
gians could wrongly be classified by the victim of attack rape as non-
Western, and Non-Western men are more likely to be police-reported
than rapists of Norwegian/Western origin.
12,13
The OPD report under-
lines that the emphasis on rapists of non-Norwegian/non-European
origin results from a strong focus on this topic in the media in recent
years.
13
It is important that future research aims at looking behind the
surface of the immigrant over-representation of sexual offenders, in
search for nuanced and diverse explanations, which can contribute to
diminish rather than increase stereotypical misconceptions.
4.1. Limitations
A limitation of the study is that data were collected retrospectively
and partly by self-reporting. It is likely that some information about for
example threats or physical violence has been lost as amnesia due to
intoxication and/or psychological stress is common among rape vic-
tims. Unfortunately, for some of the variables from both data sources,
there is a considerable amount of missing information. The information
regarding police investigations are limited in this study, us not being
allowed to look into police logs or notes, only to the official police
records (BL). Especially disappointing was that education and em-
ployment data of identified suspects were not available in the records.
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
6
This could have given us a better picture of the suspects.
Our data represent only one of the 27 police districts in Norway,
which limits the national generalizability of the findings. There is no
reason to believe that the suspect pool in rape and attempted rape looks
very different in the 27 police districts. For some of the variables,
however, for example the representation of suspects with a non-
Western origin, differences could be expected between urban and rural
populations. In our study only 21% were recidivists, this in contrast to
what has been found in a Danish study which reported up to two
thirds.
3
This could partly be explained by us only being allowed to
collect criminal register data from the STPD, and did not have access to
the national register of accused offending. Estimates of recidivism rates
is complicated partly due to the problem of under-reporting, and the
U.S. Department of Justice concluded in 2015 that there is universal
agreement in the scientific community that the observed recidivism
rates of sex offenders are underestimates of actual reoffending.
5
This
indicates that the percentage of recidivists among the reported suspects
in our study may also be too low. It is reasonable to believe that many
cases of first-time assailants are not reported to the police.
The analyses regarding police investigations are limited in this
study, and some of our statements may be interpreted as discredit to-
wards law enforcement and their investigational approach in cases of
sexual assault. We are aware that judicial and investigational processes
are extremely complex and demanding in this field, and that priorities
made by the police are often a question of finances.
Finally, like most research regarding rape, this study lacks suspect
information from all the cases that are never reported. Unreported adult
male rapists constitute a large population and we have no knowledge
on how they differ from assailants who are reported. Strengths of the
study are the large number of variables, the long observation period of
7½ years and a relatively large sample. The merging of data from police
files and corresponding SAC records is unique by enriching the variety
of data available, especially regarding victim sociodemographic data,
injuries and biological trace evidence. It also provides important insight
in the impact of forensic evidence collected at a SAC, on the in-
vestigation of sexual assault cases.
4.2. Conclusions
Comparing separate groups of sexual assault suspects based on as-
sault characteristics, has disclosed some obvious patterns of differences
between the groups, which can provide valuable information for future
prevention programs, police investigation strategies, health care
guidelines and for new research projects. The recidivist suspects appear
to represent a generally more vulnerable group than the first-time
suspects, in our detection of a high unemployment rate and a possible
drug abuse problem. The police seem to be more thorough in their
investigations of recidivist suspects than in the other two groups of
suspects, and recidivist suspect cases more often end with prosecution.
This seems unfair, considering that the others may be just as guilty. The
low reporting rates of sexual assault imply that most assaults go un-
recorded. It is important to report sexual assault to the police and to
health care. When more of these crimes are publicly reported, more
cases will possibly be taken to court if investigated thoroughly. This will
send a clear message to victims that reporting leads to more prosecution
and conviction. It will also send a message to potential assailants, that
rape is taken seriously, and that the risk of getting caught is increasing.
Acknowledgments
Thanks to deceased Senior Researcher and Psychologist Jim Aage
Nøttestad for making this project possible and for comments which
were always inspiring and cheerful. Thanks to Berit Marianne Bjelkåsen
at the Unit of Applied Clinical Research at the Norwegian University of
Science and Technology who has kindly provided the web-based re-
gistration form and converted it to an SPSS-file. Thanks to the students
Katharina Frydendal Pedersen, Karen Løhre, and Anna Hjorth-Hansen
who participated in the registration of the hospital data, and to the
students Amalie Steinsbekk, Amanda Rygvold, Ingvild Johansen, and
Eirik Alberto Brattheim for reading and recording the police data.
Thanks to Camilla Forr for registering forensic genetic laboratory data.
References
1. Offences investigated. Social Conditions, Welfare and Crime. 2014; 2014. Oslo:
Statistics Norway; 2016 cited 2017 December. Available from: https://www.ssb.no/
en/sosiale-forhold-og-kriminalitet/statistikker/lovbrudde/aar/2015-10-29?fane=
tabell&sort=nummer&tabell=242571.
2. Kripos. Taktisk Etterforskningsavdeling [Central Unit for Norwegian Police Investigation]
Voldtektssituasjonen I Norge 2015 [Sexual Assault in Norway, a Report on the Situation
2016; 2016 2015.
3. Nielsen LH, Hansen M, Bramsen RH, Hansen OI. Rape suspects: predicting charges
and convictions in Danish rape cases. A Coordinated Community Response to Rape
and Sexual Assault (PhD thesis). Department of psychology, Faculty of Health Sciences,
University of Southern Denmark. 2018; 2018:129–156.
4. Ingemann-Hansen O, Brink O, Sabroe S, Sorensen V, Charles AV. Legal aspects of
sexual violence-Does forensic evidence make a difference? Forensic Sci Int.
2008;180(2-3):98–104.
5. U.S. Department of Justice Office of Justice Programs. Sex Offender Management
Assessment and Planning Initiative (SOMAPI). Ch. 3: Sex Offender Typologies, Ch. 5:
Adult Sex Offender Recidivism, Ch. 6: Sex Offender Risk Assessment. 2015; 2015.
6. Hanson RK, Morton-Bourgon KE. The accuracy of recidivism risk assessments for
sexual offenders: a meta-analysis of 118 prediction studies. Psychol Assess.
2009;21(1):1–21.
7. Lussier P, Blokland A. A developmental life-course perspective of juvenile and adult
sexual offending. In: Sanders T, ed. The Oxford Handbook of Sex Offences and Sex
Offenders. Oxford University Press; 2017:241–269.
8. Maletzky BM. Sexual assault. In: Van Hasselt VB, Hersen M, eds. Aggression and
Violence. An Introductory Text. Boston: Allyn and Bacon; 2000.
9. Miethe TD, Olson J, Mitchell O. Specialization and persistence in the arrest histories
of sex offenders. J Res Crime Delinquen. 2006;43(3):204–229.
10. Kruse AE, Strandmoen JF, Skjørten K. Menn Som Har Begått Voldtekt - En
Kunnskapsstatus [Men Who Have Committed Rape] NKVTS Rapport 1 2013; 2013
([Report by Norwegian Competence Services on Violence and Traumatic Stress]).
11. Lussier P, Proulx J, LeBlanc M. Criminal propensity, deviant sexual interests and
criminal activity of sexual aggressors against women: a comparison of models.
Criminology. 2005;43:247–279.
12. Mjøs K. Er Det Noe Som Kjennetegner Voldtektsforbrytere? [Is There Anything Which
Caracterizes Rapists? Master thesis from Norwegian University of Science and
Technology 2011; 2011
13. Sætre M, Grytdal V. Voldtekt I Den Globale Byen - Endringer I Anmeldte Voldtekter Og
Seksualkultur I Oslo [Rape in the Global City - Changes in Police-Reported Rape and the
Culture of Sexual Behaviour in Oslo]. 2011; 2011 Oslo Politidistrikt [Oslo Police
District].
14. Edwards KM, Turchik JA, Dardis CM, Reynolds N, Gidycz CA. Rape myths: history,
individual and institutional-level presence, and implications for change. Sex Roles.
2011;65(11-12):761–773.
15. Smith M, Wilkes N, Bouffard LA. Rape myth adherence among campus law en-
forcement officers. Crim Justice Behav. 2016;43(4):539–556.
16. Maddox L, Lee D, Barker C. The impact of psychological consequences of rape on
rape case attrition: the police perspective. J Police Crim Psychol. 2012;27(1):33–34.
17. Campbell R, Johnson CR. Police officers' perceptions of rape - is there consistency
between state law and individual beliefs? J Interpers Violence. 1997;12(2):255–274.
18. Shaw J, Campbell R, Cain D, Feeney H. Beyond surveys and scales: how rape myths
manifest in sexual assault police records. Psychol Violence. 2017;7(4):602–614.
19. Hohl K, Stanko EA. Complaints of rape and the criminal justice system: fresh evi-
dence on the attrition problem in England and Wales. Eur J Criminol.
2015;12(3):324–341.
20. Statistisk Sentralbyrå [Statistics Norway]. 2010; 2010. cited 2018 May. Available from:
https://www.ssb.no/folketall.
21. Hagemann CT, Stene LE, Myhre AK, Ormstad K, Schei B. Impact of medico-legal
findings on charge filing in cases of rape in adult women. Acta Obstet Gynecol Scand.
2011;90(11):1218–1224.
22. Stene LE, Ormstad K, Schei B. Implementation of medical examination and forensic
analyses in the investigation of sexual assaults against adult women: a retrospective
study of police files and medical journals. Forensic Sci Int. 2010;199(1-3):79–84.
23. The General Civil Penal Code (Ch. 19, P. 76-82) Oslo Norwegian Ministry of Justice.
2006; 2006. Available from: http://www.ub.uio.no/ujur/ulovdata/lov-19020522-
010-eng.pdf.
24. Nesvold H, Ormstad K, Friis S. To be used or not to be used, that is the question: legal
use of forensic and clinical information collected in a self-referral sexual assault
centre. J Forensic Sci. 2011;56(5):1156–1162.
25. McGregor MJ, Du Mont J, Myhr TL. Sexual assault forensic medical examination: is
evidence related to successful prosecution? Ann Emerg Med. 2002;39(6):639–647.
26. Finstad L. Politiblikket [From a Law Enforcement Viewpoint. third ed. 2013; 2013 Oslo,
Norway.
27. Moller AS, Backstrom T, Sondergaard HP, Helstrom L. Patterns of injury and reported
violence depending on relationship to assailant in female Swedish sexual assault
victims. J Interpers Violence. 2012;27(16):3131–3148.
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
7
28. Hakkanen H, Puolakka P, Santtila P. Crime scene actions and offender characteristics
in arsons. Leg Criminol Psychol. 2004;9:197–214.
29. Pape H. Sexual assault while too intoxicated to resist: a general population study of
Norwegian teenage girls. BMC Public Health. 2014;14:406.
30. McWhorter SK, Stander VA, Merrill LL, Thomsen CJ, Milner JS. Reports of rape re-
perpetration by newly enlisted male navy personnel. Violence Vict.
2009;24(2):204–218.
31. Woods L, Porter L. Examining the relationship between sexual offenders and their
victims: interpersonal differences between stranger and non-stranger sexual offences.
J Sex Aggress. 2008;14(1):61–75.
32. Justisdepartementet [The Norwegian Ministry of Justice]. Fra Ord Til Handling -
Bekjempelse Av Voldtekt Krever Handling [Prevention of Rape Requires Action]. Oslo,
Norway. 2008; 2008 (Author).
B.F. Vik, et al. Journal of Forensic and Legal Medicine 71 (2020) 101771
8
