Criminal Justice Outcomes in Intimate and Non-intimate Partner Homicide Cases


3.4 Dependent variables: Decision-making stages in the criminal process

Sentencing is considered by many to be the most important stage in the criminal justice process. However, research has begun to question the validity of studies that focus exclusively on sentencing decisions and ignore pre-sentence decision-making stages (Bernstein et al., 1977a, 1977b, 1979; Hagan, 1974; Miethe & Moore, 1986; Petersilia, 1983; Schur, 1971; Swigert & Farrell, 1977). A primary concern has been the failure to consider the possible effects of earlier decisions on later outcomes (Greenberg, 1977). For example, an accused person who pleads guilty in exchange for a charge reduction may receive a lighter sentence than an accused person who does not plead guilty, but is subsequently found guilty at trial and sentenced accordingly. As a result, studies that do not examine whether a case was resolved at trial or by guilty plea may not capture the effect of this earlier decision on the length of sentence ultimately imposed. In addition, decisions made at earlier stages of the criminal process should be examined as important decision-making stages in and of themselves because the sequential nature of the criminal process means different criminal justice actors make decisions at different stages. For instance, prosecutors are initially responsible for deciding whether a case is resolved by trial or by guilty plea. The specific concerns of a prosecutor at this stage of the process may be substantively different than judicial concerns at sentencing. Therefore, examining the types of cases that reach conviction, the types of accused persons convicted by trial, or the types of cases resolved by guilty plea are all interesting questions worthy of examination.

In response to these issues, this study focuses on five key stages in the criminal court process that capture eight different decisions. The outcomes are: (1) initial prosecution charge; (2) mode of conviction: (3) verdict at trial; (4) type of acquittal; (5) overall conviction; (6) severity of conviction; (7) type of sentence; and, finally, (8) length of sentence.[12] Table 3.2 shows the coding and frequency distributions for all eight outcomes that are discussed separately below.

Table 3.2: Coding and Frequency Distributions for Eight Criminal Justice Outcomes, Total Sample, Toronto, Ontario, 1974-2002 (N=1,137) a
Variable Coding Frequency Percentage
Y1 Initial prosecution charge
0 Other charges 718 63
1 First-degree murder charge 419 37
Y2 Mode of conviction
0 Guilty plea 476 42
1 Case sent to trial 654 58
Y3 Verdict at trial
0 Acquitted at trial 259 40
1 Found guilty at trial 395 60
Y4 Type of acquittal
0 Found 'not guilty' 163 63
1 Fount 'not criminally responsible' 96 37
Y5 Overall conviction
0 Acquitted 271 24
1 Convicted 866 76
Y6 Severity of conviction
0 Other conviction 525 61
1 Murder conviction 341 39
Y7 Type of sentence
0 Provincial (2 years less one day) 151 17
1 Federal (2 years plus one day) 715 83
Y8 Length of Sentence Interval level 9.20 years

a As shown in parentheses for each outcome, there is variation in the sample size at various stages of the process because some accused persons may be selected out of the process depending on the outcome at the previous stage.

Initial prosecution charge

The first dependent variable captures the severity of the prosecution charge. This is an important decision point for an accused because the seriousness of the initial charge, under most circumstances, will be an important determinant of the ultimate sentence received (see Brereton & Casper, 1981-82). That is, the penalty structure available to the sentencing judge is contingent upon the conviction charge and the seriousness of the conviction charge is indirectly contingent on the seriousness of the initial charge (Brerton & Casper, 1981-82). For example, a first-degree murder charge may be reduced to second-degree murder upon conviction, but an accused charged with second-degree murder cannot be convicted of first-degree murder at trial. Moreover, the amount of harm caused by an accused is itself an important consideration in sentencing decisions and this is reflected, in part, by the initial charge laid (Brerton & Casper, 1981-82). However, some studies of police and prosecutorial decision-making suggest that charging practices in homicide cases may reflect a number of subjective, political considerations (Higgingbottom & Zamble, 1988; Mather, 1979; Williams & Rodeheaver, 1991). That is, police and prosecutors may 'charge up' to demonstrate that they are not being too lenient, to allow room for possible plea negotiations, or to allow for the possibility that evidence may be uncovered during the investigation that supports a more serious offence such as evidence of malice aforethought and/or premeditation

The data used in this analysis show that first- or second-degree murder charges were initially laid in 91 percent of the cases. As a result, because the majority of charges are for murder, a dichotomous measure captures the severity of the charge laid by the prosecution, distinguishing between those charged with first-degree murder – the most serious charge possible in cases of homicide – and those charged with less serious offences, including primarily second-degree murder. As discussed, above, the primary justification for a first-degree murder charge is the presence of planning and deliberation (Grant et al., 1998). Table 3.2 indicates that first-degree murder charges were laid in 37 percent of the Toronto cases. Another 54 percent of accused persons were charged with second-degree murder and approximately eight percent were charged with manslaughter.

Mode of conviction

All accused persons have a constitutionally guaranteed right to a trial by judge or jury. However, most criminal charges in Canada are disposed of by a guilty plea rather than at trial (Ruby, 1999). While guilty pleas may not be as frequent in homicide cases as they are for other types of violent or non-violent crime, they do represent a significant proportion of cases that result in convictions. A guilty plea represents an admission by the accused that he or she committed the offence they are charged with and that they consent to a conviction being entered without a trial. A plea of guilty can occur in two ways. First, the prosecution may charge an accused with a particular offence related to the homicide and the accused pleads guilty to that offence, often with the expectation of a reduced sentence. On the other hand, as part of a plea bargain, the accused may plead not guilty to the offence he or she is initially charged with, but guilty to a lesser charge. For example, an accused charged with first-degree murder may plead not guilty to this offence, but guilty to the lesser offence of second-degree murder. Again, it may often be the case that an accused hopes to reduce the number of years of imprisonment that may be imposed before parole eligibility by pleading guilty. To capture mode of conviction, a dichotomous measure distinguishes between cases that proceeded to trial and those that were resolved through a guilty plea. Table 3.2 shows that 58 percent of the accused persons in this sample had their case resolved at trial.[13] The remaining cases (42 percent) were resolved through guilty pleas.

Verdict at trial

Some research has shown that when cases of violent crime are resolved by trial, judges are less likely to convict if the accused allegedly victimized a stranger whereas juries are more likely to convict if the crime involved strangers (Myers, 1981). Beyond this, research is sparse on how the victim-accused relationship may affect an accused person’s verdict at trial.[14] To examine this outcome, a variable measures whether the accused was found guilty or acquitted at trial. In this sample, 60 percent of the cases that were sent to trial resulted in a verdict of ‘guilty’ for the accused.

Type of acquittal

There are two ways that an accused may be acquitted: By receiving a ‘not guilty’ verdict at trial or by being found ‘not criminally responsible by reason of mental disorder.’ In the first scenario, an accused is acquitted of the charge(s) before the court. With respect to the second type of acquittal, in the vast majority of cases, the accused admits that he or she committed the crime but, because they did not understand the nature and quality of their act, they are found ‘not criminally responsible by reason of mental disorder’ (prior to 1992, the verdict was ‘not guilty by reason of insanity’).[15] To examine whether those persons who killed intimate partners were more likely to be acquitted due to mental disorder, a dependent variable distinguishes between accused persons who were found ‘not guilty’ and those who were found ‘not criminally responsible.’ In the Toronto data, 37 percent of those acquitted were found ‘not criminally responsible by reason of mental disorder.’

Overall conviction

Regardless of whether a case is resolved at trial or by guilty plea, there may be variation in the overall likelihood of conviction. The fifth dependent variable represents an overall measure that distinguishes between those who were convicted for their crimes (including those convicted by trial and by guilty plea) and those who were acquitted (including those found not criminally responsible). In the Toronto data, three-quarters (76 percent) of those accused were convicted for their crimes.

Severity of conviction

An accused person’s conviction should demonstrate the degree of harm or damage caused by the crime as well as their perceived culpability. The distinction between a murder and a manslaughter conviction is of great significance for an accused in Canada because murder carries a mandatory penalty of life imprisonment whereas manslaughter convictions carry no minimum mandatory sentence (Grant et al., 1998). Generally, as already noted, murder is distinguished from manslaughter by the existence of a specific intent to cause death or serious bodily harm likely to cause death (s.229 CCC) on the part of the accused. Thus, the presence of certain extraneous or mitigating factors such as provocation (see specifically s.232(2) CCC), intoxication or diminished responsibility may reduce an offence from murder to manslaughter. To determine whether intimacy affects the severity of conviction, a variable measures whether the accused was convicted of murder (either first- or second-degree) or convicted of a less serious charge (primarily manslaughter). Table 3.2 shows that 39 percent of the cases resulted in a conviction of either first- or second-degree murder.[16]

Type of sentence

Prior literature suggests that the sentencing of an offender involves two separate decisions. The first is whether to imprison the offender and, if imprisonment is imposed, the second decision is what sentence length is appropriate for the crime. With respect to type of sentence, because homicide is considered to be a serious offence, very few cases that lead to a conviction do not result in a term of imprisonment. However, in Canada, an offender can be sentenced either to a provincial term of imprisonment (two years less a day) or to a federal term of imprisonment (two years or more). To determine whether there is an association between the type of victim-accused relationship and the type of sentence, this dependent variable captures whether an offender was sentenced to a provincial or federal institution. In the Toronto sample, as expected given the seriousness of the offence, 83 percent of the offenders were sentenced to a federal term of imprisonment.

Length of sentence

The length of time an offender is sentenced to serve before parole eligibility is based, in part, on materials presented to the court after guilt has been determined (e.g. pre-sentence reports that describe, for example, their criminal history). Sentencing decisions are also dependent, however, on the type of conviction imposed and, within each offence category, the range of possible sentences that are available. For example, offenders convicted of first-degree murder in Canada receive an automatic life sentence with no chance of parole for 25 years – the longest possible term of imprisonment for a homicide in this country. In contrast, while a second-degree murder conviction also carries a mandatory life sentence, the period of parole ineligibility may range from 10 to 25 years. Finally, there is no minimum mandatory sentence for manslaughter convictions. Thus, the final dependent variable is a continuous measure that captures the number of years an offender has been sentenced to serve before parole eligibility with values ranging from zero up to and including 25 years. During the study period, the average term of imprisonment before parole eligibility in the Toronto court was approximately nine years. Of those convicted of manslaughter, the average sentence was 5.5 years and, of those convicted of second-degree murder for which the minimum mandatory sentence is 10 years, the average sentence was 12.5 years. All those convicted of first-degree murder are sentenced to 25 years before parole eligibility.

Date modified: