-
BRIT.J. CRIMINOL. VOL 38 NO. 3 SUMMER 1998
PUBLIC PERCEPTIONS OF HOMICIDEAND CRIMINAL JUSTICE
BARRY MITCHELL*
This article offers an account of a recent quantitative surveyi
of public opinion on homicide and thecriminal justice system. The
survey set out to test the assumptions made in substantive criminal
lawthat the public recognize variations in moral culpability
between homicides and that thesevariations should be reflected in
separate offences carrying separate sentencing provisions. If
thisfundamental assumption is well founded, the survey further
sought to determine what factorsappear to influence the public's
assessment of gravity in homicides, how the law should
categorizehomicides, and what justifications and excuses should be
recognized. Finally, it set out to test publicopinion as to the
appropriate penalties for the most serious homicides.
There is a growing body of research evidence of the impact of
public opinion aboutcrime seriousness and sentencing policy both in
this country (Levi and Jones 1985) andelsewhere (Golash and Lynch
1995), and policy makers are undoubtedly influenced byperceptions
of public sentiment and the need to maintain public confidence in
diecriminal justice system (Indermaur 1991; Roberts 1992). At a
theoretical level, it hasbeen forcefully argued that public opinion
has a crucial role to play in denouncingcriminal behaviour and
reaffirming society's values (Davies 1993).
The vast majority of criminological research has investigated
attitudes across a widerange of offences. Various polls have
elicited views on capital punishment, but apartfrom two snapshot
surveys by the MORI organization in October 1990 (for the
BBCprogramme Public Eye) and in August 19952 (for the News of the
World newspaper),there is a conspicuous dearth of detailed
investigation of opinion on homicide and howthe criminal justice
system should respond to it. The public appear to have anambivalent
interest in homicide; for some time there has been a demand
forpublications depicting detailed accounts of sensational and
usually horrific killings,whilst those same crimes have been
roundly condemned, especially by the press andother sections of the
media.
Smith's (1984) study based on the inhabitants of a small area in
Birmingham foundthat for 52 per cent of her sample the main source
of information about crime was themedia,3 and this is supported by
research in other jurisdictions (e.g. Knowles 1982;Broadhurst and
Indermaur 1982; and Canadian Sentencing Commission 1987).Research
also suggests that (i) the news media usually provide insufficient
informationon which to make reasoned evaluations of events, so that
opinions are formed without
School of International Studies and Law, Coventry University,
Coventry.1 This survey was funded by two grants; one for 55,900
from the Leverhulme Trust, the other for f 15,700 from Coventry
University's Research Initiative.1 A quota sample of 1,527
adults aged 18 years or more were interviewed on 4 August 1995 in
the street at 68 constituency
sampling points throughout Great Britain. The data were weighted
to match the profile of the population.9 35.8 per cent relied on
hearsay or the supposed experience of friends and neighbours, 3.2
per cent used their own experience,
and just 1.3 per cent used the police.
453
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
any substantive knowledge of the issues (Hans and Slater 1983);
(ii) the mediainvariably reports only the traditional punitive
responses to crime (Barrile 1984;Roberts and Doob 1990); (iii) the
media's apparent influence on attitudes towardscrime, through
programmes such as Crimewatch UK, Crime Stoppers andCrimebusters,
has lead to a fear of crime (Home Office 1989) and (iv) the 'true
crime'realm provides an inaccurate picture and understanding of
homicide (Durham et al.1995). At the same time, Sparks (1992) warns
us against assuming that televisionnecessarily influences the
public's perception of crimefictional crime programmesare watched
primarily for entertainment rather than enlightenment.
Although the law recognizes very specific kinds of homicides
through crimes such asassisting suicide and infanticide, the
majority of criminal killings are categorized aseither murder or
manslaughter, with the former offence representing what areregarded
as the most serious homicides. The criminal law has long recognized
adistinction between murder and manslaughter (Kaye 1967), and the
decision to retainseparate offences rather than a single crime of,
say, 'criminal homicide', is based largelyon the hitherto untested
assumption that it accords with public opinion (Criminal
LawRevision Committee 1980, paras. 15, 23 and 28; Goff 1988).
It is further assumed that, in broad terms, public opinion on
the issue mirrors thecriminal law in its adoption of the principle
of fair labelling, that the law should reflectdegrees of wrongdoing
through distinct offences, separately labelled and
frequentlycarrying distinct sentencing provisions, so that
offenders can be formally stigmatized inproportion to their
blameworthiness (Criminal Law Revision Committee 1976, para.7). The
details of the conviction should indicate the degree of harm or
damage and theoffender's moral culpability, so that the label
attached to the crime and the criminalrepresents a proportionate
response to unlawful conduct (see, for example, Ashworth1981;
Glanville Williams 1983; Horder 1994). Abolition of the murder/
manslaughterdistinction would, it is argued, lead people to think
that the law had been weakened andthat 'the law no longer regarded
the intentional taking of another's life as beingespecially grave'
(CLRC 1980, para. 15). Furthermore, the singling out of murder
asrepresenting what is commonly believed to be the worst homicides
is also used to justifythe imposition of a mandatory sentence of
life imprisonment.4
Yet some of our most senior judges have publicly argued that in
its current formmurder does not always represent the most heinous
form of homicide (see, e.g. LordKilbrandon in Hyam v. DPP,5 and
Lord Hailsham in R v. Howe et al.6). Moreover, itseems that in
practice the verdicts in homicide cases, whether dealt with as
guilty pleasor as contested trials, do not necessarily reflect the
theoretical differences betweenmurder and manslaughter (Mitchell
1991). This apparent discrepancy could not
* In the more recent of the two MORI surveys (i.e. the one
conducted in August 1993) referred to above, 71 per cent said thai
rapeof a child and 65 per cent thought that rape of an adult should
justify the death penalty, and these figures are very similar to
thepercentages of respondents who favoured capital punishment for
certain types of murder.
5 'It is no longer true, if it was ever true, to say that murder
as we now define it is necessarily the most heinous example of
unlawful
homicide' [1974] 2 WLR 607 at 640.s 'Murder, as every
practitioner of the law knews, though often described as one of the
utmost heinousness, b not in (act
necessarily so, but consists in a wliole bundle of offences of
vastly differing degrees of culpability, ranging from brutal,
cynical andrepeated offences like the so-called Moors murders to
the almost venial, if objectively unmoral, "mercy killing" of a
belovedpartner1 [1987] 2 WLR 568 at 581.
454
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
simply be attributed to evidential issues but may well result
from the personal views ofthose involved, both lawyers and
jurors.
Assumptions about the state of public opinion have also been
made in therationalization of the law's approach to justifications
and excuses in homicide. In caseswhere battered spouses kill their
abusers, the law has been criticized for its narrowinterpretation
of reacting with 'a sudden and temporary loss of self-control',
oftenthereby depriving the defendant of a manslaughter conviction
based on provocation.Wasik (1982), for example, argued for 'less
emphasis upon the need for an immediatereaction against gross
provocation' and urged the law to recognize 'shades ofculpability,
even in deliberate killings', claiming that 'all the evidence there
is points tothe public's opposition to stigmatising killing under
cumulative provocation asmurder'.
In practice, mercy killing is not infrequently treated as
manslaughter by straining theconcept of diminished responsibility
so as to produce what is perceived as a more justoutcome, namely a
manslaughter conviction which in turn does not force the court
toimpose a life sentence (CLRC 1976; House of Lords Select
Committee 1989). The basisof this is that '[n]o-one connected with
[a mercy killing] case wants to see the defendantconvicted of
murder' (CLRC 1980). A person who kills in self-defence
theoretically hasa defence even to a murder charge provided it was
reasonable for him to use force in thecircumstances and no more
than a reasonable degree of force was used; if he exceedsthe bounds
of reasonableness by even the smallest amount, he is likely to be
convicted ofmurderthe law adopts an 'all-or-nothing' approach (e.g.
R v. Clegg [1995]). Onceagain, though, this is a subject about
which lawyers seem to think they have their fingerson the public
pulse. In Palmer v. R [1971], one of the leading authorities on the
point,Lord Morris, commented, 'If a jury thought that in a moment
of unexpected anguish aperson attacked had only done what he
honestly and instinctively thought was necessarythat would be most
potent evidence that only reasonable defensive action had
beentaken'.7 It is difficult not to harbour a suspicion that
ordinary people would at leastreduce the conviction of a killer who
used excessive force in self-defence.
Furthermore, the decision of the House of Lords in R v. Howe et
al. [ 1987] to excludeduress by threats (of death or serious
injury) as any kind of defence to murder ispresumed to be broadly
in line with public acceptance. Lord Hailsham remarked, 'Ihave
known in my own lifetime of too many acts of heroism by ordinary
human beings ofno more than ordinary fortitude to regard a law as
either "just or humane" whichwithdraws the protection of the
criminal law from the innocent victim and casts thecloak of its
protection upon the coward and the poltroon in the name of a
"concession tofrailty'". However, commentators have been very
critical of this 'heroic argument' as arationale of the law (Dennis
1987; Milgate 1988; and Walters 1988). Indeed,commenting on the
ruling in Howe, Professor Smith (1987) claimed, 'Even the
mostresolute and well-disposed citizen would be likely to comply
(with the threat)'.
7 It is worth noting here that until quite recently the
Australian courts regarded homicides resulting from the use of
excessive
force in self-defence as manslaughter, but this was reversed by
the High Court in Ztctvic v. DPP (1987) 61 ALJR 375 so that
suchcases are treated as murder. In the Republic of Ireland
however, the Supreme Court has held that a person who uses no more
forcethan lie honestly thought was necessary but objectively did
use excessive force should be guilty of manslaughter, see PtopU
(A-G)v.Dvjrr [1972] IR 416.
455
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
The Study
In collaboration with Social and Community Planning Research
(SCPR), a survey wasundertaken with the primary aim of testing the
law's assumption that it has publicsupport for the way in which it
deals with homicides. To do this we set out to examinepublic
opinion in England and Wales on a range of specific homicide
scenarios, onhomicides generally, and how the criminal justice
system ought to respond to them.More specifically, the research
sought:
(a) to determine whether the public recognize variations in
gravity betweenhomicides;
(b) if they do,(i) to identify factors which they treat as
influencing the gravity of homicides;(ii) to determine whether they
feel such variations should be reflected by the
legal recognition of separate offences carrying separate
sentencingprovisions; and
(c) to determine what kinds of sentencing powers the public
think ought to be avail-able to the courts when dealing with
defendants convicted of'criminal homicides'.
Before embarking upon the main national survey, two pilot
studies were undertaken.The first of these, carried out with the
assistance of NOP/MAI and funded by a smallgrant from the Nuffield
Foundation, was in the form of moderated group discussions.
Itconfirmed that the public recognize the concepts of murder and
(albeit slightly lessreadily) manslaughter, though not exactly as
they are construed in law, and it appearedto suggest that the
public do recognize that some homicides are more serious
thanothersi.e. that they merit different levels of punishment and
condemnation. Afurther pilot study tested a draft questionnaire and
confirmed the feasibility of thegeneral approach of using a series
of homicide scenarios as part of a
quantitative,interviewer-administered, questionnaire survey. The
survey was based on a randomsample designed to represent the
general population in England and Wales, drawnfrom the Postcode
Address File.8 The response rate was 71 per cent, yielding
822interviews. On key demographic variables, the sample reflected
the population aged 18or over in England and Wales.
Respondents were asked to consider the following eight
hypothetical scenarios:Card A: Burglary (hereafter the 'burglary
scenario')A burglar was disturbed by the owner of the house, a
25-year-old woman. He panicked.He picked up an ashtray which was
near at hand and hit her over the head with it. Shedied of her
injuries.Card B: Terminally ill woman (hereafter the 'mercy killing
scenario')A woman was terminally ill and in great pain. She had
been begging her husband to 'puther out of her misery' for months.
Eventually, he gave in to her request and suffocatedher whilst she
was asleep.
1 The interview! were conducted in October 1995.
456
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
Card C: Woman drowning (hereafter the 'drowning woman
scenario')A young woman was walking her dog by a lakeshe slipped
and fell into the water. Apasser-by saw her drowning in the lake.
Although he could swim, instead of trying tosave her, he walked by
and she drowned.Card D: Man and woman arguing (hereafter the 'thin
skulls scenario')A man and woman were arguing about who was first
in the queue at the supermarket.He gently pushed her, and she
unexpectedly tripped and bumped her head against awall. She had an
unusually thin skull and she died from her injuries.Card E:
Mountain climbers (hereafter the 'necessity scenario')Two mountain
climbers were roped together. One of them slipped and fell. The
othertried to hold on to the rocks for both of them, but he knew
that if he did not cut the rope,they would both die. To save
himself, he cut himself loose, knowing that the otherclimber would
fall to his death.Card F: Battered spouse (hereafter the 'battered
spouse scenario')A woman had been physically and sexually abused by
her husband for three years. Hecame home one evening and started
hitting her again. She felt that she couldn't standany more abuse,
so she waited until her husband was sleeping, then hit him over
thehead with a saucepan, killing him.Card G: Two men arguing at
work (hereafter the 'self-defence scenario')Two men were having a
heated argument at work which developed into a Gght. One ofthem
picked up a screwdriver and lunged at the other. Fearing that he
would otherwisebe stabbed, the unarmed man grabbed a spanner, and
in self-defence he hit the otherman over the head with it, killing
him.Card H: Killer threatened with his own life (hereafter the
'duress scenario')A group of terrorists threatened a man with his
own life if he did not agree to kill a localbusinessman. The man
was told by the terrorists that he had a week in which to kill
thebusinessman, and that if he went to the police he would be
shot.9 Scared for his own life,the man could see no alternative and
he killed the businessman.The law currently treats these homicide
scenarios very differently both in terms ofoffence category and
sentencing. Respondents were asked to
(1) rank them in order of seriousness (i.e. those which were the
most blameworthyand deserved the greatest punishment), and rate
them on a 1 to 20 scale (with 20representing the most serious
killing), giving their reasons for the rating;
(2) suggest appropriate sentences for the highest and lowest
rated scenarios; and(3) respondents were invited to reconsider some
scenarios in which one factor had
been altered, in order to test the potential relevance of
further factors.
0 Tlie killer in this scenario was given a fairly long period in
which to kill the businessman for two reasons: (i) to accentuate
the
apparent difference between this case and scenario E which
raises the possible defence of necessity or 'duress of
circumstancej'; and (ii)because the law on the degree of immediacy
of execution of the threat in duress per minas is unclear. The
law's ultimate concern seemsto be wheilicr the accused's will was
overpowered ahen cotnuiUmg Iht offence (see Lord Edmund-Davies in
DPP for N Inland v. Lynch[1975] AC 653 at 700), although a threat
of future violence might persuade the court that D killed because
of the threat hanging overhim rather than because of the immediate
pressure on him; tee Lord Parker CJ in R v. Hudson and Taylor [1971
] 2 QB 202 at 206,207.
457
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
Respondents were also asked to consider homicides generally and
to describe what theyregarded as die worst homicides and suggest
how they should be sentenced, and toidentify what they saw as the
least serious killings.
The Homicide Scenarios
(a) Ranking the scenariosWhether the average gravity of the
eight scenarios is calculated by averagingrespondents' rank orders,
or by averaging their gravity scale ratings, the same
rankingemerged. This is shown in Table 1 below.
TABLE 1 Ratings of homicide scenarios on 1-20 scale
Mean Standard deviation
Card A; burglary10 15.5 4.0Card H: killer threatened with own
life" 12.7 4.7Card C: woman drowning12 9.8 5.4Card G: two men
arguing15 9.1 4.4Card F: battered spouse" 8.1 5.5Card D: woman with
thin skull15 5.9 4.3Card E: mountain climbers16 5.6 4.2Card B:
terminally ill woman" 3.5 4.1
10 Sixty-nine per cent of Uie sample rated card A at 15 or more.
There were no significant differences in the rating of card A
(burglary scenario) by respondents' age, gender, qualificatiom
or religion, but there were some noteworthy variations based
onsocial class and marital status. Those in skilled manual
occupations rated the case slightly more seriously (16.29) than
those in othersocial daises; and those living with a partner tended
to rate it less severely (14.61) dian others.
11 Very few variations were found between die ratings of
different sub-groups for card H (the duress scenario).
"Women rated card C (the drowning woman scenario) more severely
than men (mean rating of 10.2 compared to 9.34), as didyounger
(18-24 year-olds) and older respondents (65 or over) compared to
the intermediate age groups (mean ratings 11.22,10.66and 9.26
respectively). Those with children rated the case less severely dun
those widiout (9.55, compared to 10.49).
"Women again rated card G (die self-defence scenario) more
severely than men (mean rating 9.62 compared to 8.46); so too
didolder (75 or over) respondents (10.11 compared to 8.79 for die
under-35s).
H Older (die over-65s) and younger (the under-25s) respondents
rated it more severely than other age groups (mean ratings
9.93, 8.5 and 7.57 respectively). Of particular interest,
married, cohabiting or divorced/separated respondents thought it
wasdistinctly less serious than those who were single and never
married or widowed (7.8,6.98 and 7.29 compared to 9.01 and
9.94respectively). Additionally, those with a religion and those
with no qualifications (mean ratings 8.49 and 8.96) regarded it
moreseverely than those who had no religion and those wiui
educational qualifications (7.27 and 7.78).
18 Fifty-eight per cent of respondents gave this scenario no
more dian 5 out of 20. Respondents living with a partner rated card
D
(the thin skull scenario) significantly lower than those who
were widowed, divorced/separated or single (means ratings of 5.79
and4.44 compared to 6.1, 6.54 and 6.28).
'"Sixty per cent gave uiis scenario no more than 5 out of 20.
Women rated card E (die necessity scenario) more severely dian
men(6.19 compared to 4.98), as did younger (under 25) and older (65
or over) respondents compared to those in intervening agegroups
(6.43 and 6.73 compared to 5.15). Widowed and single respondents
thought this case was more serious than their marriedcounterparts
(6.34 and 6.1 respectively compared to 5.31).
17 77.8 per cent rated diis scenario at 4 or less. Single
respondents who had never inarried tended to give a slightly more
severe
rating to card B (mercy killing scenario) than their married or
cohabiting counterparts (4.29 compared to 3.19 and
2.50respectively). Those with a religion also gave it a higher
rating (3.84) Uian uiose who did not have one (2.56), as did
respondentswith no qualifications (4.24) compared to those widi
qualifications (3.13). Similarly, respondents at the two ends of
the age spectrumthought card B merited a higher score (4.51 and
4.61 from 18-24 year olds and 65 years or more respectively) than
those between25 ajid 64 (2.90).
458
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
Trie burglary scenario was perceived as the most serious
homicide scenario by 53 percent of the respondents. Three-quarters
placed it in the first two, and almost nine out often put it within
the three worst killings. Respondents under 25 and those aged at
least65 years were less likely to regard this as the most serious
of the scenarios. Those withchildren, however, were more likely
than those without to rate it as one of the two mostheinous
homicides (77 per cent compared to 67 per cent).
The most common reason for giving the burglary killing a high
seriousness scorewas the fact that the killer was already
committing a crime. More than four in ten saidthat he should not
have been in the victim's home in the first place, and roughly
threein ten reasoned he had set out with the intention of
committing a crime (albeitburglary rather than murder or
manslaughter). Thirteen per cent also thought thefact that the
deceased was killed in her own home aggravated the matter.
Othercriticisms of the killer in card A were: he should have run
away when confronted; thekilling was premeditated,I8 there was no
reason for the killing; the killer's motive wassimply to save
himself, and he must have intended to injure and was willing to use
aweapon. Interestingly, those who regarded this scenario as
relatively seriouselsewhere in the survey described their
perception of'the worst kinds of homicide' byreference to the
vulnerability of the victimthe victim was innocent or
defencelessetc.
Almost three-quarters of the sample placed the duress scenario
in the three worsthomicides. However, the rating of this scenario
should be treated with caution since itappeared that some
respondents were assessing the severity of the threat per se to
thekiller rather than his motives and justification. There were no
significant differencesbetween respondent subgroups. The most
common reasons for regarding the scenarioas comparatively serious
implied that the killing was avoidable. Four out often thoughtthe
killer could have gone to the police, and a further four in ten
said he had analternative course of action. Just over a fifth
regarded the fact that the killing wasplanned or calculated as the
most aggravating feature, and about one in seven criticisedthe
killer's motivesaving himself. Others pointed to the cowardly
nature of the killingand to the victim's innocence.
The drowning woman scenario was generally placed third of the
eight, though itprovoked a wide range of responses. Forty-four per
cent placed it in the worst three, 335placed it fourth or fifth,
and almost a quarter put it in the three least serious
scenarios.Younger respondents, especially those under 34 years,
tended to regard it morecritically. The scenario most frequently
(four in ten) received the comment that thekiller should have tried
to save her. Other aggravating features were the disregardshown for
human life, behaving contrary to the normal rules of society, and
the fact thatthe killer had an alternative course of action.
like the drowning woman scenario the battered spouse scenario
provoked a broadrange of reactions. Almost a third rated it as one
of the three worst cases, four out oftenthought it one of the three
least serious, and 27 per cent perceived it as of middlingseverity.
Interestingly, there were very few significant differences in the
responses ofsubgroups. The severity rating varied considerably
within each subgroupage,
" The contrary was wggested in the (acti.
459
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
gender, social class etc. The only statistically significant
variation was by marital status;married or cohabiting respondents
were slightly less likely than single or'never-married' respondents
to rate this scenario as one of the two most serioushomicides (14
per cent compared to 25 per cent).
Many of the respondents who were relatively critical of the
battered spouse scenarioalso identified the presence of an
alternative course of action as a particularlyaggravating
characteristic. Others noted the fact that the killing was
premeditated, andearlier in the survey those who treated this
scenario as serious were more likely todescribe the worst homicides
by reference to the killer's motive. Those who rated thebattered
spouse scenario (card F) as of relatively low severity said the
killer had been'driven to the end of her tether'. Others said she
needed to save her own life, or that shehad suffered such abuse.
Earlier in the survey these respondents, when describing theiridea
of the worst homicides, placed more stress than other respondents
on the victim'svulnerability and less on the method of killing or
the victim's suffering: when describingthe least serious homicides,
they again put more emphasis on the victim's vulnerability,and less
on the killing being accidental.
Fifty-five per cent of respondents placed the thin skulls
scenario in the three leastserious homicides, and there were no
significant differences between subgroups. It wasrated lowly
largely because the killing was perceived as accidentalmore than 60
percent said the killer bore no fault, the killing was
unpremeditated or death wasunforeseeable. Those who thought this
was a relatively less serious homicide were morelikely than those
who regarded it more gravely generally to characterize the
worsthomicides by reference to the killer's motive.
The necessity scenario (card E) was regarded as one of the three
least serioushomicides by 63 per cent of the sample and, again,
there were no significant differencesbetween subgroups. Nearly a
quarter felt the killer had no alternative, or at least neededto
save his own life. One in ten said it was not the killer's fault;
slightly fewer thought thevictim must have known the risks
involved, and about 5 per cent of respondents saidthey would have
done what the killer did in the same circumstances. Respondents
whogave this a low gravity rating were less likely than other
respondents to describe the leastserious homicides as 'accidental',
but more likely to identify the killer's motive as afactor in the
worst homicides.
The mercy killing scenario was perceived by more than half the
sample as the leastserious of the eight scenarios, as one of the
two least serious by 72 per cent, and as one ofthe three least
serious by 83 per cent of respondents. About three-quarters of
women,those with higher qualifications and those without a religion
placed this as one of thetwo least serious scenarios, compared to
about two-thirds of men, those with fewerqualifications and those
with a religion. It was relatively 'well' received primarily
(i.e.four out of ten respondents) because the killing was 'for the
right reasons' or an act ofmercy. Other common reactions were: the
victim wanted to die, she had a poor qualityof life and was in
great pain, people should have the right to die with dignity, and
shewould have died anyway. Interestingly, when asked elsewhere in
the survey for a generaldescription of the most serious homicides,
those who rated this scenario very low on thegravity scale were
more likely than those giving it a higher rating to focus on the
killer'smotive. Yet they were less likely than other respondents to
talk about the killer's motivewhen describing the least serious
homicides, and more likely to focus on the
victim'svulnerability.
460
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
(b) Sentencing the highest and lowest rated scenariosThe views
of respondents as to what they regarded as the appropriate
sentences for thescenarios which they rated as either the worst or
the least serious of the eight are shownin Tables 2 and 3 below.
(They were not specifically asked to indicate whether
theirresponses referred to die nominal sentence or the time
actually served in prison,though some expressed the view that there
should be no difference between die two.)
About one in six respondents recommended the death penalty for
the burglary andduress killers, whereas just over half of the
respondents thought that a determinateprison sentence of at least
ten years would be appropriate. The sentences for the killersin the
other scenarios in Table 2 were noticeably less severe. Obviously,
it waspredictable that respondents would take a generally
sympathetic approach towardssentencing in scenarios given the
lowest rating (Table 3), diough it is worth noting thatthe
self-defence scenario attracted some recommended short prison
sentences, albeitthat the figures are clearly very small.
(c) Scenario variationsWhen asked to re-assess the burglary
scenario (card A) on the basis that the victim wasaged 70 radier
than 25, two-thirds of the respondents thought that the
severityremained unchanged, whereas 30 per cent regarded the
killing as more serious. At diesame time, 52 per cent diought the
punishment should be tougher and 47 per cent saidit should be no
different. The proportion of respondents who felt that die victim's
agemade a difference decreased widi the level of educational
qualifications: 41 per cent ofthose widi no qualifications regarded
killing an older person as more serious comparedto only 26 per cent
of those with degree-level qualifications.
Fifty-six per cent of die sample diought diat the killing would
be more serious if diedrowning woman (card C) had been die man's
daughter dian if she had been a complete
T A B L E 2 Sentence recommended for scenario given the highest
severity rating
DeathNatural life imprisonment30 yrs + imprisonment20-29 yrs
imprisonment15-19 yrs imprisonment10-14 yrs imprisonment5-9 yrs
imprisonmentLess than 5 yrs imprisonmentNon-custodial
sentencePsychiatric counsellingOther penaltyNo penaltyCan't say
Burglaryscenarion=473(%)
16201214151551_
-
1-
1
Duressscenarion=163(%)
161910141413651-
-
11
Drowningwomanscenarion=84(%)
21459
20221326188
Batteredspousescenario:n=54(%)
138-
-
151720_
16-
-
2
Self-defence, mercykilling, necessity andthin skull
scenariosn=50(%)
1415858
275412-
64
461
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
TABLE 3 Sentence recommended for scenario given lowest severity
rating
Mercykillingn=496{%)
Thin skullscenarion=144(%)
Necessityscenarion=124(%)
Batteredspousescenarion=53(%)
Drowning woman.self-defence, burglaryand duress
scenariosn=86(%)
Death _ _ _ _ _Natural life imprisonment - - 2 - 230 yrs +
imprisonment - - - - -20-29 yrs imprisonment - - - 1 115-19 yrs
imprisonment - 2 - - 510-14 yrs imprisonment - - - - 15-9 yrs
imprisonment 1 2 2 1 1 0Less than 5 yrs imprisonment 7 26 6 10
20Non-custodial sentence 5 23 10 9 5Psychiatric counselling 32 15
20 43 22Other penalty 2 1 2 1 3No penalty 51 27 51 23 21Can't say 2
6 6 13 12
stranger.19 Fifty-seven per cent said that the punishment ought
to be tougher. Lookingat subgroupings, 63 per cent of those with no
qualifications regarded letting a daughterdrown was more serious
than letting a stranger drown, compared to 50 per cent of thosewith
degree-level qualifications.
If the battered spouse (card F) killed her husband without any
premeditation,two-thirds of the sample thought the offence would be
less serious, with a further 30 percent saying it made no
difference. Four-fifths thought the punishment should be
lesssevere. No variations between subgroups could be detected.
Homicide generally
(a) The worst homicidesOnly 29 respondents chose one of the
scenarios as representing the type of homicide towhich they would
give a rating of 20, the worst possible type of killing.20 Fourteen
chosecard A (the burglary scenario), ten chose card H (the duress
scenario), two each chosecards F and B (the battered spouse and
mercy killing scenarios), and one chose card D(the thin skull
scenario).
In describing their idea of the worst possible homicide,
respondents tended to focuson specific aspects of the killing; most
commonly (71 per cent) the type of victim, butalso (67 per cent)
the killer's apparent primary motive, and (38 per cent) the method
ofkilling. Ten per cent of the sample referred to an actual
homicide recently reported inthe media.
10 Forty-one per cent aid it should make no difference.
Furthermore, the killing of a dose relative u likely to lead to
tougher
penalties in, for example, Italy, Switzerland and Turkey.1 0
Five respondent! felt that all killings were of equal gravity
and were thus unable to describe a 'worst possible' homicide.
462
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
Fifty-five per cent of the sample felt that a 20-rated homicide
involved a childvictim;21 other descriptions spoke of a
'defenceless victim' (37 per cent), an 'innocentvictim' (20 per
cent), or an 'elderly victim' or policeman/fireman (9 per cent
each).Women were more likely than men to choose child killings as
the worst (61 per centcompared to 48 per cent), as did respondents
with children or grandchildren (59 percent compared to 44 per cent
of those without). The reasons most commonly given forchoosing
child victims as representing the worst homicides were the
innocence ordefencelessness of the deceased and the fact that
children were deprived of a long lifeexpectancy (23 per cent, 58
per cent and 14 per cent respectively).
The next most commonly cited feature of the worst homicide
concerned the killer'smotive, or lack of motive. A third of
respondents regarded sexually motivated killingsas relevant here;
27 per cent mentioned premeditated or 'cold-blooded' killings,
and17 per cent referred to apparendy motiveless homicides. Those
from higher socialclasses were distincdy more likely to explain
dieir choice of 'die worst homicide' interms of die killer's
motive. This may be linked to die higher level of
educationalqualificadons57 per cent of respondents widi no
qualificadons mendoned modvescompared to 72 per cent of those widi
qualificadons.
As to die relevance of die mediod of killing, about one in eight
(13 per cent) talked ofhorrible or violent homicides, and 16 per
cent mentioned cases in which die victim issubjected to torture or
long suffering before dying. Abduction before deadi, especiallyof
child victims, was mentioned by 7 per cent of the sample.
(b) Sentencing the worst homicidesThe views of respondents as to
what they regarded as the appropriate sentences for dieeight
scenarios and for what diey treated as the worst possible homicides
are set out inTable 4 below.
TABLE 4 Sentences recommended for worst of scenarios and for
worst possible homicides
Recommended sentence Worst of 8 scenarios (%) Wont possible
homicide (%)Death sentence 13 55Natural life imprisonment 17 34At
least 30 years'imprisonment 10 220-29 years' imprisonment 12 415-19
years'imprisonment 13 110-14 years' imprisonment 16 15-0 years'
imprisonment 8 134 years' imprisonment 3 -1-2 years' imprisonment
2Other term of imprisonment 1 1Psychiatric treatment and/or
counselling 2 -Other punishment 1 2No punishment 1 -Can't say/no
response 3 2
51 In Uie 1905 MORI survey ihe highest level of support (i.e. 82
per cent) for the death penalty was for die inurder of a child.
463
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
It is important to acknowledge that it was not possible to test
the extent to which theserespondents agreed about the nature of
what constitutes the worst homicides.Nonetheless, men were more
likely than women to propose the death penalty (61 percent compared
to 50 per cent), and slightly less likely than women to recommend
anindefinite prison sentence (27 per cent compared to 39 per cent).
Capital punishmentwas also more likely to be desired by those with
fewer qualifications, and those from thelower social classes. There
was no apparent significance in the sentencingrecommendations made
by respondents with a religion and those without, althoughthere
were some variations between different religions. Sixty per cent of
the Church ofEngland Christians favoured the death penalty,
compared to 39 per cent of the RomanCatholics and 50 per cent of
other Christians.
(c) The least serious homicidesHomicides to which respondents
attributed a 1 rating largely fell into three
maincategorieseuthanasia or mercy killing (49 per cent), accidental
killings (30 per cent),and killing in self-defence or
self-preservation (13 per cent). Seven out of tenrespondents chose
one of the eight scenarios as illustrative of the least serious
homicide.Forty-three per cent chose card B (the mercy killing
scenario), and 14 per cent pickedout card D (the thin skulls
scenario).
There were slight variations between subgroups in rating
euthanasia as the leastserious homicide. Women, parents and those
with educational qualifications were morelikely than men, those
without children and respondents with no qualifications toidentify
euthanasia here (55 per cent, 52 per cent and 51 per cent compared
to 44 percent, 41 per cent and 42 per cent respectively). Of those
with a religion, RomanCatholics were less likely than other
Christians to treat such homicides as the leastserious (38 per cent
compared to 50 per cent of Church of England Christians). In astudy
in which the fieldwork was carried out in the spring of 1995 (i.e.
just a few monthsbefore the homicide study was undertaken), the
authors of British Social AttitudesSurvey (hereafter the BSAS)
concluded that 'in certain limited circumstances, there
isoverwhelming public support for euthanasiafor the patient in an
irreversible comaon a life-support machine (if their relatives
agree) and for people with incurable andpainful illnesses who have
requested euthanasia', and 'strong support for the notionthat
patients should be able to rely on doctors to end their lives on
the basis of writteninstructions, such as tfiose in "living wills'"
(Jowell et al. 1996). For the BSAS, though,statistically
significant factors were qualifications, disability, ethnicity and
religiousbeliefs, so that the homicide survey only partially
reiterates the BSAS findings.
The reasons for regarding euthanasia in this way focused
primarily on the victim'sfeelings or situation: 37 per cent spoke
of the victim's wish to die, 23 per cent of the rightto die, and 26
per cent cited the victim's poor quality of life and die fact of
his/hersuffering.22 In addition, though, 42 per cent identified the
killer's good motivekillingfor the 'right reasons', to relieve
suffering, made it one of the least serious homicides.
"Six per cent of the sample said they would want to die in those
circumstances, and 3 per cent said they would have done the sameas
the killer did.
464
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
Accidental killings were seen as amongst the least serious
because there is no fault onthe part of the killer, because there
is no intent to kill, and because the killer could nothave foreseen
the outcome of his actions. This latter observation by respondents
isparticularly interesting in the light of the law's view that a
person's foresight of theconsequences of his action is evidence of
his intention.23 Women and parents wereslightly less likely than
men and those with no children to regard accidental homicidesas the
least serious (26 per cent and 28 per cent compared to 35 per cent
and 36 per centrespectively), as were Roman Catholics compared to
other Christians (41 per centcompared to 29 per cent).
Just over one in eight respondents rated self-defence or
self-preservation homicidesas the least serious. Most of those (83
per cent) referred to the killer's motiveheneeded to save his own
life, or he had no alternative course of action. Some said
thekilling was not deliberate or premeditated.
Offence categories (groups of scenarios)One of the principal
aims of the survey was to determine whether the public
recognizevariations in the gravity of homicides and, if they do,
whether they feel such variationsshould be reflected in convictions
for different offences such as murder andmanslaughter, or some
alternative form of formal labelling. To this end, respondentswere
asked to place the eight scenarios into different piles, with each
pile representing aseparate category of crime.
The 811 respondents who completed this task created piles that
made up 129different combinations of the scenarios. On average they
created 4.4 piles, and men andthose with no educational
qualifications tended to make slightly fewer piles than womenand
those with such qualifications. Almost eight out often respondents
felt that at leastone scenario represented a distinct type of crime
and thus placed it in a 'pile' of its own.Half the sample treated
card B (mercy killing scenario) in this way, and about a
thirdsingled out cards A (burglary scenario) and C (drowning woman
scenario) as distinctoffences.
In piles made up of two or more scenarios, the most common
associations (44 percent and 37 per cent respectively) were between
cards D (the thin skull scenario) and G(self-defence scenario), and
cards A (burglary) and H (duress). To a slightly lesserextent
respondents associated cards D (thin skull) and E (necessity) (31
per cent), andcards C (drowning woman) and E (necessity) (30 per
cent).
It is obviously very difficult to state with any confidence what
the outcome of thescenarios would be from a purely theoretical
point of view, for the facts given are soscant. However, it might
well be suggested that cards A, E and H would probably resultin
murder convictions, and that B and D would lead to manslaughter
verdicts. Theoriginal scenario F (battered spouse) theoretically
looks like murder though in practicea manslaughter verdict would
not be uncommon. Card C (drowning woman) is nocrime, unless the man
is related to the woman (as in the varied version), in which
casethe verdict should potentially be murder. Card G (self-defence
scenario) is either no
" Though whether a person intends a consequence depends on
wliether he foresees it with sufficient degree of certainty
orprobability.
465
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
crime (if the killer's use of force was reasonable and
proportionate), or murder (if heacted unreasonably).
It was therefore interesting to find a good deal of linkage
between scenarios A and H(37 per cent of respondents) and to a
slightly lesser extent between E and H (21 percent), but much less
so between A and E (only 8 per cent). Scenarios B and F were
placedin the same category quite frequently (26 per cent), though
fewer than one in five linkedB and D or D and F (17 per cent and 14
per cent respectively).
As for piles of three or more scenarios, the strongest
associations (13 per cent) werebetween cards C, D and E and between
D, E and G.
As indicated in the introduction, the law assumes that the
public wishes to reflect itsviews about the varying degrees of
gravity in homicides through the separate offences ofmurder and
manslaughter. This labelling of crimes is also a means of
encapsulating theessence of the wrongdoing and simultaneously
manifesting fundamental distinctionsbetween different categories of
offending.24 Thus, respondents were asked to give aname or
description to each category they had created. Predictably, the
names theysuggested varied considerably according to the scenarios
in the category.
Half the respondents used the word 'murder' to denote categories
linked with card A(the burglary homicide), and just over two-fifths
used it to describe the categories linkedwith card H (the duress
killing).25 In contrast, only about 7 per cent referred
tocategories linked with cards B, D or E (the terminally ill woman,
the thin skull woman,and the mountain climbers respectively).
The word 'killing' was used by about 15 per cent of respondents
to describe categoriescontaining cards A, C, F, G and H (the
burglary, the drowning woman, the batteredspouse, the self-defence
killing, and the duress scenario), by 16 per cent to
describecategories containing card E (the necessity killing), and
by 19 per cent for those linkedwith card D (the thin skull
scenario). One in ten used it for categories linked with card B(the
mercy killing).
'Manslaughter' was chosen by roughly 15 per cent of respondents
to describecategories associated with cards D (the thin skull
scenario) and G (the self-defencekilling), and by about 10 per cent
to describe homicides containing cards A (theburglary), H (the
duress scenario) and card B (the mercy killing).
Another quite commonly used term was 'accident' (or
'accidental'). More thanfour-fifths of those who placed card D
(thin skull) in a category by itself described it inthis way; so
too did more than half the respondents who linked card D with
otherscenarios. 'Accident' or 'accidental' also described 18 per
cent of homicides linked withcard G (self-defence), 17 per cent for
those linked with card E (necessity), and 12 percent for those
linked with card C (drowning woman).
'Self-preservation' was chosen to depict homicides associated
with cards G(self-defence), E(necessity), F(batteredspouse)and H
(duress)by 21 percent, 20percent, 14 per cent and 11 per cent
respectively.
** Tluu, the law not only distinguishes murder from manslaughter
but also voluntary manslaughter (which usually denotes all
theelements of murder together with some mitigating feature such as
provocation), from mvohtnlary manslaughter (where the killing isnot
'wiili malice aforethought').
u Almost two-thirds (64 per cent) of respondents who grouped
cards A and H in the same offence category described that
category as 'murder'. So too did more than a third of those
Unking cards H and C (the drowning woman).
466
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
The only other relatively frequendy used names were
'premeditated' or 'planned',which described homicides linked widi
cards H (duress scenario), A (burglary), and cardF (battered
spouse)by 15 per cent, 12 per cent and 12 per cent
respectively.
Obviously, one of the arguably unique features of card C (woman
drowning) concernsthe issue of guilt through doing nodiing
(omission to act). It is worth noting that of the34 per cent of
respondents who regarded this scenario as a 'category of its own',
9 percent described it as 'omission', 17 per cent as 'indifferent'
or 'not caring', 10 per cent as'neglect', 7 per cent as
'negligence' and another 7 per cent as 'cowardice'.
Implications and Conclusions
It is always difficult to draw confident conclusions from a
quantitative survey in theabsence of any more detailed qualitative
follow-up research. Nonetheless, this surveyprovides strong
evidence that the initial assumption made by the criminal law that
diepublic recognize variations in the gravity of homicides is well
founded. Furthermore,the survey clearly suggests diat die public
would like these variations to be reflected inthe legal recognition
of separate offences.
What is also encouraging for the law is that superficially at
least die factors which diepublic identify as influencing die
gravity of homicides broadly mirror its approach tothe issue, to
the extent that gravity appears to be made up of two principal
components,namely die harm caused26 and die offender's personal
culpability. However, oneimmediate reservation here is diat die law
differentiates between gravity in homicidessolely on die basis of
die killer's culpability, apparendy assuming that eidier die
harmelement remains constant for all cases or diat any variation in
levels of harm isinsufficiendy significant to justify separate
offence categorization.
Aldiough die majority of die factors respondents identified as
affecting dieirassessment of gravity in homicide essentially
concerned culpability, diere wereoccasions on which diey referred
to some aspect of harmfulness. For example, there wasdie response
from a number of people diat homicides involving young victims
arerelatively serious because diere is a considerable loss of life
expectancy. Thesignificance of harm variations was reinforced here
by die further response thathomicides in which die victim is
tortured or endures lengthy suffering are particularlyserious.27
Conversely, one of die reasons for perceiving die mercy killing
scenario asamongst die least serious was die deceased's poor
quality of life: in crude terms, beingkilled did not deprive the
deceased of any significant right or interest.
Personal culpability, diough, is predictably and undeniably an
important factor indie public's evaluation of homicides.
Comparatively serious homicides were said to becharacterized by
premeditation28 or planning or a victim who is defenceless
orodierwise vulnerable. It was interesting to find that one of die
criticisms of die passer-bywho made no attempt to save die drowning
woman scenario (card C) was diat his actions
1 0 Unfortunately, there was insufficient opportunity in the
survey to test, for example, public opinion on offence
categorization
and sentencing where there b more than one deceased.17
Germany, Italy, Spain and Switzerland also look upon homicides
where the victim suffered cruelty as particularly serious.M
Various other jurisdictions also treat premeditation or planning
as aggravating features in homicide; see, for example,
Italy,Luxembourg, Spain, Switzerland and Turkey.
467
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
manifested a disregard for human life. This focus on an
individual's attitude towardsharm or injury has been advocated as
indicating one of the ways in which the law'straditional narrow
emphasis on cognition through its construction of mens rea shouldbe
reformed.29
Accident or lack of fault, or justifiable self-preservation (as
where there is noreasonable alternative course of action) were
associated with the least serioushomicides. However, one dimension
of culpability which the law largely seeks to avoidwhen framing
offences and identifying justifications and excuses concerns the
killer'smotive, and yet this very factor featured quite prominently
in respondents'assessments.50 Sexually motivated killings, or
apparently motiveless killings wereregarded as amongst the most
serious, whereas killing for 'good reasons' such as therelief of
suffering or self-preservation rendered the homicide much more
acceptable.Significant numbers of the public appear to have some
concept of good and badmotives, but whether they have sufficiently
similar concepts, whether law reformers candraft suitable
definitions, and whether adequate safeguards can be created to
ensurethat the law would not then be abused is not at all dear.
As well as identifying factors relating to harm and culpability,
respondents alsoreferred to matters of wider social policy. For
example, mercy killing was frequendygiven a low gravity rating on
the basis that people ought to be able to exercise a right todie.
(Again, there remains uncertainty about the adequacy of the level
of support forthis and concern about the danger of abuse before
there can be any serious suggestionto alter the law.) Conversely,
the drowning woman scenario was often regardedrelatively seriously
because the passer-by showed a disregard for human life;
moreover,he could easily have made some sort of attempted rescue,
and his failure to make anyattempt went 'against the normal rules
of society'. Yet it should be acknowledged thatthese responses were
given with the benefit of hindsightrespondents were told thatthe
passer-by could swim and that the woman drownedand it is difficult
to see howthis kind of censure could be satisfactorily accommodated
by the law. For although itmay only imply there should be a general
duty to take reasonable measures to try toprevent fatalitieson the
basis that the prevention of death is sufficiently differentfrom
the prevention of other forms of harm as to justify an
exceptionthere areobvious problems. It may not be entirely clear at
the relevant time whether death orsome lesser form of injury is
likely, nor might the probability of death be apparent.Thus, it is
not easy to see how a court could make die judgment in any given
set of factsthat the defendant was duty-bound to attempt a rescue.
Hitherto at least, the criminallaw has only imposed a duty to take
positive action by virtue of an unambiguous set ofpre-existing
circumstances.31
The study strongly suggests that ordinary people agree with the
principle of fairlabelling, at least in relation to homicide, that
the nature and magnitude of diewrongdoing should be reflected by
die recognition of distinct offence categories.
" S e c , for example, Duff(1990) and my 'Culpably Indifferent
Murder' (1996: 64-86). It b also worth noting that Dutch lawregards
D'I indifference as to whether liii actions will kill another as
indicative of an intent to do so.
3 0 Motive is quite commonly treated as an aggravating feature
in other jurisdictions; see, for example, Germany, Italy,
Spain,
Switzerland and Turkey.91
Such as the relationship (e.g. parent/child) between die
parties, some contractual obligation on the defendant vis-a-vis
thevictim, or where the defendant has created the danger.
468
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
Unfortunately, though not unpredictably, die ways in which
respondents placed diescenario cards into different piles did not
generally provide any really dear indicationof how homicides might
be categorized or subgrouped. The statistics did not suggestany
obviously striking associations between groups of scenarios, and
there wasinsufficient time available during the survey in which to
ask respondents what preciselythe names of dieir piles signified.
One of the potential complications here was dieuncertainty about
the way in which ordinary people would use die distinction
betweenthe nature of die convicdon (murder or manslaughter etc.)
and die sentence, to reflectvariations in gravity.
Aldiough diere was insufficient time in which expressly to
elicit respondents' views onprecisely where to draw die line
between criminal and non-criminal homicides, someclues may be found
in dieir comments about sentencing. The fact diat
significantnumbers of respondents recommended no punishment for die
killers in die mercykilling, the diin skulls scenario and die
necessity scenario obviously implies a perceivedlack of moral
culpability at least. The ratings and recommended sentences for
otherhomicides involving self-defence, duress, battered spouses and
omitting to try to savelife were more equivocal, making it unwise
to offer any implications about justificationsor excuses for
killing.
It is difficult to draw any precise conclusions about die
public's assessment of mercykilling. That 43 per cent should rate
it one out of 20 when advised diat die scale rangesfrom die least
to die most serious homicide and reminded diat diey do not have to
usedie extreme numbers on die scale, obviously suggests diat
significant numbers ofpeople regard it as not amounting to a crime.
At the same time diough, whilst diemedian rating was 1.12, die mean
was 3.45. Of die 403 respondents who placed card Bin a category by
itself, 80 per cent labelled it 'mercy killing' or 'eudianasia',
whereas only3 per cent described it as 'not a crime' or 'not a
killing', and whilst this dearly does notnecessarily imply an
element of wrongfulness, nor does it automatically signal
completelegitimacy or acceptance.
The diin skulls scenario (card D) also received a 1 rating from
a number ofrespondents, and diis represents die kind of homicide
which die Law Commission(1996) recendy recommended should cease to
be treated as meriting a conviction formanslaughter, largely on the
ground that die killer lacks suffident moral culpability.This type
of homidde was rated generaUy slighdy higher than die mercy
killingscenario, and was scored very similarly to die case of the
mountain climbers (card E)where the essential issue was whether the
circumstances adequately justified or excusedthe killing. Bearing
in mind diat diere was dien a gap to die ratings of die next group
ofscenarioscards C (drowning woman), F (battered spouse), and G
(self-defencescenario)it may be diat respondents would not regard
the diin skulls and mountaincumbers homicides as crimes but would
regard higher rated scenarios as offences.Obviously, diough, diis
is a matter which requires furdier examination.
Furthermore, whilst die law distinguishes between duress (by
direats or bycircumstances) and self-defence as adequate
justification or excuse for homidde, diedifferentiation recognized
by die public appears to lead to radier different
condusions.Mention has already been made of die unfortunate way in
which respondents in thissurvey approached their evaluation of card
H (die duress scenario) in which die defencemight be based on
duress by threat of death. Nevertheless, whatever form
ofmeasurement is used, die mountain dimbers scenario (card E) was
regarded as
469
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
distinctly less serious than the two men arguing scenario (card
G) where self-defencewas appropriateand yet however much sympathy
might be felt the law would treat themountain climber as a
murderer, whereas the killer in self-defence would be acquitted
ifthe court thought he had acted reasonably and had used
proportionate force against hisvictim. Moreover, given the low
rating of the necessity scenario, it may be that ifpresented with a
homicide in which there is clearly no fair opportunity to adopt a
thirdcourse of action, significant numbers of respondents would be
very sympathetic towardsa person who killed under duress by
threats.
The burglar who killed when disturbed by his victim (card A) was
undeniably andpredictably treated, both in relative and absolute
terms, as a very serious offender. Thatthis was a 'crime within a
crime' was dearly an aggravating feature and, as such, itresembles
the concept of'felony murder' which is widely regarded in the USA
as one ofthe more serious forms of homicide, notwithstanding the
fact that it is an example ofconstructive liability.52 However,
respondents on average rated this scenario at" 15 to 16out of a
possible maximum of 20, and only 14 people chose it as representing
their ideaof the worst homicide. This surely suggests that for many
members of the public thereare other, distinctly more serious,
forms of homicide.
The point was earlier made that although respondents seemed to
regard thedrowning woman scenario (card C) quite seriously, with an
average rating of between 9and 10 out of 20, there was no
opportunity to investigate the extent to which they havethought
through the implications of extending the existing law requiring
people totake positive action in specific situations. At the same
time, it may well be that this ratingreflects the traditional view
of'the sanctity of life' and that taking life is a unique form
ofharm (Fletcher 1978), though this does not necessarily mean it is
the most seriousharm.33
It was also quite predictable that respondents would want to see
tougher penalties forthose who kill. However, it was unexpected
that litde more dian half the samplefavoured any form of capital
punishment in any circumstances, and they may have haddifferent
forms of homicide in mind. For example, the British Social
Attitudes 12threport (1995) showed that in four surveys between
1983 and 1994, 70-77 per cent ofrespondents favoured capital
punishment for 'murder in the course of a terrorist act',67-71 per
cent for 'murder of a police officer' and 59-66 per cent for 'other
murders'.The current survey suggests diat the British Social
Attitudes authors' comment thatpublic support for the
re-introduction of the death penalty seemed to be 'at anirreducible
minimum' is perhaps unduly pessimistic! Respondents in this latest
surveyhad already spent some time thinking about different homicide
situations reflectingvarying degrees of seriousness, so diat when
asked to comment on sentencing diey werevery aware that some
homicides do contain elements of mitigation. More specificquestions
about punishment appear to attract more moderate responses.
Nevertheless,an aggregate figure of just 55 per cent supporting the
death penalty in anycircumstances is surprisingly low.
5 1 i.e. the accused may neither intend to kill nor knowingly
risk killing another perion, but the law holds him responsible
and
liable Tor killing because he ought not to have been committing
a crime anyway. Other jurisdictions, such as Germany, Luxembourgand
Turkey, also view killing in the furtherance of crime as especially
serious.
u A n interesting analysis of the harm in homicide can be found
in Joel Feinberg (1984: 79-83), where die author rehearsescompeting
arguments about what amounts to 'the most serious harm'.
470
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
PUBLIC PERCEPTIONS OF HOMICIDE AND CRIMINAL JUSTICE
The discovery that about a third of respondents favoured natural
life imprisonmentfor homicides appears to support Hough's (1996)
conclusions that although the publicoften advocates types of
sentence which are in line with judicial practice, for
seriouscrimes they tend to have punitive views about sentence
length.*4 Large numbers ofrespondents in the survey preferred
either a much longer determinate period ofimprisonment or a life
sentence in the literal sense. It is commonly supposed that
thepublic are bemused by the idea that life sentence prisoners
should ever be released onlicence; to many, it is simply a
contradiction in terms.
REFERENCES
ASHWORTH, A. J. (1981), The Elasticity of Mens Rea', in C.
Tapper, ed., Crime, Proof andPunishmentEssays in Honour of Sir
Rupert Cross. London: Butteiworths.
BARRILE, L. (1984), Television and Attitudes about Crime', in R.
Surette,/astic and the Media.Springfield, IL: Thomas.
BROADHURST, R. and INDERMAUR, D. (1982), 'Crime Seriousness
Ratings: The Relationship ofInformation Accuracy and General
Attitudes in Western Australia', Australian and New Zealand
Journal of Criminology, 15: 219-34.CANADIAN SENTENCING
COMMISSION (1987), Sentencing Reform: A Canadian Approach.
Ottawa:
Ministry of Supply and Services.CRIMINAL LAW REVISION COMMITTEE
(1976), Working Paper on Offences against the Person.
London: HMSO.(1980), Fourteenth Report: Offences against the
Person. London: HMSO.DAVIES, M. (1993), Punishing
CriminalsDeveloping Community-based Intermediate Sanctions,
esp.
Ch. 1. Connecticut: Greenwood Press.DENNIS, I. (1987),
'Developments in Duress'.Journal of Criminal Law, 51: 463-80.DUFF,
R. A. (1990), Intention, Agency and Criminal Liability. Oxford:
Basil Blackwell.DURHAM, A. M., ELROD, H. P. and KINKADE, P. T.
(1995), 'Images of Crime and Justice: Murder
and the True Crime" Genre', Journal of Criminal Justice, 23/2:
143-52.FEINBERG, J. (1984), Harm To Others. New York: Oxford
University Press.FLETCHER, G. P. (1978), Rethinking Criminal Law.
Boston: Litde, Brown and Co.GOFF, R. (1988), The Mental Element in
the Crime of Murder', Law Quarterly Review, 104:
30-59.GOLASH, D. and LYNCH, J. P. (1995),'Public Opinion, Crime
Seriousness, and Sentencing Policy',
American Journal of Criminal Law, 22/3: 703-32.HANS, V. P. and
SLATER, D. (1983), 'John Hinckley Jr. and the Insanity Defense: The
Public's
Verdict', Public Opinion Quarterly, 47: 202-12.HOME OFFICE
(1989), Report of the Working Group on the Fear of Crime. London:
HMSO.HORDER, J. (1994), 'Rethinking Non-fatal Offences against die
Person', Oxford Journal of Legal
Studies, 14: 335.HOUGH, M. (1996), 'People Thinking About
Punishment', The Howard Journal, 35/3: 191-214.
M Generally, research evidence presents a slightly mixed picture
of the relationship between public opinion and judicial
practice
in sentencing. Earlier studies in England suggested a good deal
of correlation between the two (e.g. Hough and Moxon 1983). Inthe
USA, Canada and Australia, the balance of the evidence did "not
sustain the conclusion that the public is overwhelmingtjr
morepunitive than die courts' (Roberts 1992).
471
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from
-
BARRY MITCHELL
HOUGH, M. and MOXON, D. (1985), 'Dealing with offenders: public
opinion and the views ofvictims', The Howard Journal 24/2:
160-75.
HOUSE OF LORDS SELECT COMMITTEE (1989), Report on Murder and
life Imprisonment, HL Paper78-1.
INDERMAUR, D. (1991), Crime Seriousness and Sentencing: A
Comparison of Court Practice and thePerceptions of a Sample of the
Public and Judges m Western Australia. Canberra:
CriminologyResearch Council of Australia.
JOWELL, R., CURTICE, J., PARK, A., BROOK, L. andAHRENDT, D.
(1995), British Social Attitudes: the12th report. Aldershot:
Dartmouth.
JOWELL, R., CURTICE, J., PARK, A., BROOK, L. and THOMSON, K.
(1996), British Social Attitudes: the13th report. Aldershot:
Dartmouth.
KAYE, J. M. (1967), T h e Early History of Murder and
Manslaughter', Law Quarterly Review, 83:365-95,569-601.
KNOWLES, J. (1982), Ohio Citizen Attitudes Concerning Criminal
Justice, 3rd edn. Columbus, OH:Governor's Office of Criminal
Justice Services.
LAW COMMISSION (1996), Legislating the Criminal Code:
Involuntary Manslaughter, Law Com. No.237. London: HMSO.
LEVI, M. and JONES, S. (1985), 'Public and Police Perceptions of
Crime Seriousness in Englandand Wales', British Journal
ofCriminology, 25/3: 234-50.
MILGATE, H. P. (1988), 'Duress and the Criminal Law: Another
About-turn by the House ofLords', Cambridge Law Journal, 61-76.
MITCHELL, B. J. (1991), 'Distinguishing Between Murder and
Manslaughter', New Law Journal,141:935-7,969-71.
(1996), 'Culpably Indifferent Murder', Anglo-American Law
Review, 25/1: 64-86.ROBERTS, J. V. (1992), 'Public Opinion, Crime
and Criminal Justice', Crime and Justice, 16:
99-180.ROBERTS, J. V. and DOOB, A. N. (1990), 'News Media
Influences on Public Views of Sentencing',
Law and Human Behaviour, 14: 451-68.SMITH, J. C. (1987),
Commentary, Criminal Law Review, 480-5.SMITH, S. (1984), 'Crime in
the News', British Journal of Criminology, 24: 289.SPARKS , R.
(1992), Television and the Drama of Crime: Moral Tales and the
Place of Crime in Public Life.
Buckingham: Open University Press.WALTERS, L. (1988), 'Murder
under duress and judicial decision-making in the House of
Lords',
Legal Studies, 8: 61-73.WASIK, M. (1982), 'Cumulative
Provocation and Domestic Killing', Criminal Law Review,
29-37.WILLIAMS, G. (1983) 'Convictions and Fair Labelling',
Cambridge Law Journal, 85-95.
DPP for Northern Ireland v. Lynch [1975] AC 653Hyam v. DPP
[1974] 2 WLR 607Palmer v. R [1971] AC 814People (A-G) v. Dwyer [
1972] IR 416Rv.Clegg [1995] 1 AC 482R v. Howe et al. [1987]2 WLR
568R v. Hudson and Taylor [1971] 2 QB 202Zecevic v. DPP (1987) 61
AJLR 375
472
at University of N
ottingham on O
ctober 29, 2014http://bjc.oxfordjournals.org/
Dow
nloaded from