A review and estimate of time spent in prison by offenders sentenced for murder
This Web page has been archived on the Web.
Legislation guiding sentencing for murder has changed and evolved considerably over the last several decades. To examine the average length of time that offenders sentenced for murder in Canada spend in prison, it is important to consider recent and past changes in legislation in the areas of pardoning and parole.
The purpose of the present study is to determine the average time offenders sentenced for murder spend in prison over three legislatively-relevant time periods: pre-1961 (pre-capital/non-capital murder sentences); 1961 to 1976 (capital/non-capital murder designation) and 1976 to 2002 (1st and 2nd degree murder designation). Average time served in prison was examined using statistical methods commonly used for this type of data. Specifically, Kaplan-Meier survival analyses were applied to produce average length of incarceration within the above specified time periods. Length of incarceration was defined as the interval beginning with the start of the murder sentence to either of the following events: death, court ordered freedom, royal prerogative of mercy, or conditional release.
All available data for federally sentenced offenders were extracted from CSC's (Correctional Service of Canada) automated database (Offender Management System; OMS). As of April 2002 information pertaining to murder sentences was available for 4,228 offenders. This excludes inmates transferred from a foreign country and those serving murder sentences under the Young Offender's Act.
An examination of the survival curve for pre-1961 murder sentences revealed an average length of incarceration of 19.6 years. Between 1961 and 1976, the average time served in prison was found to be 15.8 years for those serving sentences for capital murder and 14.6 years for non-capital murder. This drop was most likely associated with legislative changes and the introduction of the Parole Act (1959). Historically, the most dramatic effects on the length of incarceration for offenders serving sentences for murder occur after 1976. The average incarceration time for offenders serving sentences for 1st degree murder was found to be 22.4 years, an increase of 6.6 years over sentences that fell under the capital murder definition. Note hate this is an underestimate given the maximum incarceration length of this time period is currently not observable.
Over time, the length of incarceration varied due to differing mechanisms affecting release. Prior to 1961, murder sentences were more likely to end in death or be commuted to life, resulting in a wide-ranging distribution of sentence lengths. From 1961 to 1976 commuted sentences and the new non-capital murder sentences were eligible for parole. The shortest average incarceration times are observed during this time period, particularly for capital cases. From 1976 onward, more stringent criteria have been applied to 1st degree murder sentences, extending the parole ineligibility period to 25 years, with the possibility of early parole existing at 15 years for some cases. Average incarceration lengths in this time are the most they have ever been for the more serious murder sentences, while those for 2nd degree murder sentences increased only slightly.
The authors would like to thank Colette Cousineau for her expertise in navigating Correctional Service of Canada's Offender Management System and for her work in ensuring the quality and accuracy of the data.
The Criminal Code in Canada is federal, thus, there is a single definition of murder and specification of punishment for the entire country. Legislation guiding sentencing for murder, however, has changed and evolved considerably over the last several decades. For this reason, to examine the average length of time that murderers in Canada spend incarcerated, one must first gain a clear understanding of the pertinent legislature of the past, and how changes in this legislation, particularly regarding the topics of pardoning and parole over the years, may effect such an analyses.
An investigation into the sentencing structure of the past revealed that, for federal offenders serving sentences for murder, release decisions were customarily directed via judicial discretion. Historically, offenders serving sentences for murder could be released, pardoned or have their sentences reduced through the discretionary powers of the Monarch (Evans, 1971). Eventually the concept of parole evolved and murder offenders became eligible for parole release at various stages of their sentences.
Prior to September 1, 1961, any person convicted of murder in Canada was automatically sentenced to death, with their sentence being carried into execution. At this time, however, it was possible for some prisoners to be released or have their sentences commuted to life through the royal prerogative of mercy. This type of intervention rested with the powers of the Governor General. Historical evidence indicates that the royal prerogative was exercised frequently and operated flexibly. Between Confederation (1867) and 1962, the year of Canada's last execution, the federal Cabinet commuted just under half of all death sentences to life imprisonment (Strange, 2001). Decisions to execute or spare were made on a case-by-case basis, not according to formal rules of evaluation. These commuted sentences would later be affected by future parole legislation.
Meanwhile, throughout the period of 1899 to 1959, the Ticket of Leave Act introduced the concept that release was an important part of the rehabilitative process.
"…the granting of release should be the function of a specialized body; and there should be supervision during release with penalties for violations of that release."
(Sentence management manual, 2001).
Under the terms of this act, the Governor General of Canada could grant a conditional release to any prisoner serving a term of imprisonment. Although not applied to death sentences, conditional release later became possible for those sentences commuted to life imprisonment.
On February 15, 1959, the proclamation of the Parole Act resulted in the abolition of the Ticket of Leave Act. The Parole Act resulted in a board whose members would be responsible for the granting and revocation of parole and who would make such decisions without the benefit of a face-to-face interview with the parole applicant (Sentence management manual, 2001). Parole was defined as the authority granted to an inmate to be at large during their term of imprisonment. Under the Parole Act, the Board would, at particular times prescribed by the regulations, review the case of every inmate serving a sentence of imprisonment of two years or more, whether or not an application had been made by, or on behalf of, the inmate. Inmates sentenced with murder were still only eligible for release under mechanisms such as reduced sentences, pardons, and the royal prerogative of mercy. However, this group would soon be split, allowing for certain cases to qualify for parole release.
Amendments to the Criminal Code in 1961 formally differentiated between life and death sentences. Changes resulted in murder being divided into capital and non-capital murder. With this amendment, Capital murder was defined as "murder that is planned and deliberate, murder committed in the course of certain crimes of violence by the direct intervention or upon the counseling of the accused; and murder of a police officer or prison warden, acting in the course of duty, resulting from such direct intervention or counseling." (Chandler, 1994). Such murder was still punishable by mandatory hanging, except if the accused was under eighteen years of age. All other murder, referred to as non-Capital, was punished by life imprisonment. In addition to this amendment, in 1961 an automatic review of all capital convictions by the provincial Court of Appeal was established as well as a full right of appeal to the Supreme Court of Canada. This was a review only of fact or law, in the conviction, since the sentence was mandatory and could be reduced only by the Cabinet.
As legislation changed the definition and sentence structure of murder, parole regulations impacted sentence lengths. Parole Regulation 6(b) stipulated that inmates sentenced before January 4, 1968 to life imprisonment for non-capital murder were eligible for parole after serving seven years of their sentence incarcerated. Those who received a capital sentence for murder during this time in which a sentence of death was commuted to life became eligible for parole after serving 10 years of their sentence incarcerated. Between 1968 and 1974, the parole eligibility date for non-capital murder sentences was extended to 10 years after incarceration (Parole Regulation Section 7). Consequently, for this time, the ineligibility periods for capital and non-capital murder were the same (National Parole Board, 1991).
In 1976, subsections 218(5) and (6) of the Criminal Code were amended, making the period of parole ineligibility vary between 10 and 20 years. Specifically, this legislation affected I) individuals sentenced between January 1, 1974 and July 25, 1976 to life imprisonment for a capital murder in which the sentence of death was commuted or deemed to be commuted to a sentence of life imprisonment; and ii) those individuals sentenced to life imprisonment for murder not punishable by death in proceedings that commenced during this period of time, even though a finding of guilt was made on or after July 26, 1976, were affected. These individuals also became eligible for judicial review after serving 15 years of their sentence incarcerated (National Parole Board, 1991).
The Criminal Law Amendment Act (No.2) was proclaimed to be of force and effect on July 16, 1976, formally abolishing capital punishment, as well as changing the punishment for murderers and other serious offences. This Act changed the frequency of mercy and flexibility in favor of mandatory minimum sentences for murderers (Strange, 2001). According to the Criminal Code, there are two forms of murder: first-degree and second-degree. First-degree murder is the more serious of the two and, for this reason, carries the higher penalty. First-degree murder is one that is planned and deliberate, also encompassing contract killings and the murder of police officers and prison employees. Second-degree murder entails all murder that is not first degree. Since 1976, the mandatory penalty for an individual convicted of first-degree murder is life imprisonment, with no eligibility of parole before 25 years of the sentence has been served in prison. Second degree murder also carries a mandatory penalty of life imprisonment, with a parole ineligibility period of ten years minimum, however the sentencing judge may vary that period from 10 years to a maximum of 25 years.
For first and second-degree murder sentences without parole, the offender remains imprisoned for life. Offenders who are paroled while serving life sentences must remain on parole for life unless parole is revoked. If parole is revoked, the offender is re-incarcerated. In many cases, murderers remain in prison until they die. Despite the parole ineligibility periods accompanying these mandatory life sentences, some offenders are released from prison before their original parole ineligibility period is up. When the Criminal Law Amendment Act was passed in 1976, it included a provision known as the "Faint Hope Clause". This provision refers to the process of judicial review, whereby an offender may apply to the court for a reduction in the amount of time served incarcerated before they become eligible for parole. This Judicial Review does not mean an early parole for the offender, and it does not reduce the sentence imposed by the court. This "Faint Hope Clause" simply allows certain offenders who have served at least fifteen years of a life sentence to apply to a court to have their cases reviewed by the National Parole Board at an earlier date. In order to be eligible for judicial review, the offender must have: been convicted of high treason or first degree murder (resulting in an automatic life sentence with no eligibility for parole for 25 years) or; must have been convicted of second degree murder and sentenced to life imprisonment without eligibility for parole for a period of greater than fifteen years; and have served at least fifteen years from time of arrest.
As discussed, legislation pertaining to the sentencing of murderers has fluctuated over the last several decades. This legislative disparity raises interest in the average length of time that murderers spend incarcerated. In determining the incarceration time of offenders sentenced for murder, it is imperative to consider changes that have historically influenced their sentences. This notion is reinforced in a previous calculation that estimated the average time that first-degree murder offenders spend in prison to be 28.4 years (Beavon, 1995). Beavon noted the importance of considering the Criminal Law Amendment Act as a factor impacting the average incarceration time. At the time of the analyses, however, there was not enough information to provide a reliable estimate for the post 1976 population. Thus, Beavon's calculations did not fully reflect the impact of the changes in legislation.
The purpose of the present study is to determine the average time murder offenders spend incarcerated. The report investigated this average over three distinct time periods:
- pre-1961 - pre-capital/non-capital murder sentences
- 1961 to 1976 - capital/non-capital murder designation
- 1976 to 2002 - 1st and 2nd degree murder designation
These time periods were used to reflect benchmark legislative changes that affected murder sentences. Incarceration time for offenders serving a murder sentence was defined as the time from arrest to first release or death. Depending on the legislation of the time, first release could have been a pardon, a sentence reduction or a conditional release.
Finally, the historical time line was broken down to finer intervals to examine the combined effects of legislation and the parole regulations on incarceration time for murderers.
Difficulties in estimating average time incarcerated for murder sentences arise when considering all the data available. Specifically, there will be a proportion of ‘suspended’ observations. These would comprise offenders who leave the population prior to what has been defined as a release. For example, offenders murdered during their incarceration time. Likewise, there will be a group of offenders for which there is no observable outcome to date (i.e. life sentences for which the parole ineligibility period has not been reached). These sentence outcomes are referred to as right-censored observations. Average incarceration time was examined using statistical methods commonly used for this type of data. Namely, survival techniques were applied to produce average times over the specified time intervals.
For the purposes of this research paper, all available data for federally sentenced offenders were extracted from CSC's automated database (Offender Management System; OMS). As of April 2002, information pertaining to murder sentences was available for 4,228 offenders. This excludes inmates transferred from a foreign country, and offenders serving murder sentences under the Young Offender's Act.
Of the study population, 1.9% (N = 79) were under the general definition of murder (pre-1961), 16.2% (N = 685) were designated capital/non-capital murder sentences (1961-1976) and 81.9% (N = 3,464) were either first or second-degree murder sentences (1976-2002). The majority of inmates were male (97%) with 146 (3%) women. The average age at sentencing was 30 for men and 33 for women. Note that five percent of the sample died in prison.
|Sentence Year||Murder Sentence||Total|
Kaplan-Meier survival analysis was used to calculate average incarceration time for murder sentences within the above time periods or strata. Incarceration time was defined as the interval beginning with the start of the murder sentence to either of the following events: death, court ordered freedom, royal prerogative of mercy, conditional release.
A review of legislation and an investigation into the associated effects on murder sentences suggested dividing the history timeline into three distinct strata. As anticipated, survival analyses showed varying average incarceration lengths for murderers over the three time periods.
An examination of the survival curve for pre-1961 murder sentences revealed a mean incarceration period of 19.6 years. However, a broad range of incarceration periods for this time period was observed. At the higher end, ten percent of this population was incarcerated for over 40 years. This reflects the trend in death sentences commuted to life imprisonment and, not until much later, becoming eligible for parole. Note that almost half of all death sentences during this time period were commuted to life (Strange, 2001). Conversely, 10% of all murder sentences spent under 9 years incarcerated. Possible explanations for this are; the enactment of the Parole Act (1959), reduced sentences, pardons, and sentences that ended with execution.
Between 1961 and 1976, a combination of legislative changes and the introduction of the Parole Act had a significant effect on the average incarceration time of murder sentences. For this time period, a significant drop in average incarceration was observed. Average incarceration time for capital and non-capital murder was 15.8 years and 14.6 years respectively. The combined average was 14.7 years, four and a half years less than that of the pre-1961 era. The observed drop can be attributed to a number of factors. First, the newly created National Parole Board (1959) automatically reviewed sentences for parole release, including murder sentences commuted to life imprisonment. The result was more releases of eligible cases. These cases could perhaps be equivocated to the later-defined non-capital and 2nd degree murder cases.
Next, the split of murder into two definitions, and the concept of parole eligibility made it possible for some murder sentences to qualify for release in this time period. The new capital murder definition still carried the death penalty, and non-capital murder carried a penalty of life imprisonment. Not much later, however, Parole Regulation 6(b) (1968) stipulated that non-capital murder sentences become eligible for parole after 7 years. In addition, capital murder sentences commuted to life imprisonment became eligible for parole after 10 years. Finally, Parole Regulation Section 7 (1974) extended the parole ineligibility period for non-capital murder to equal that of sentences commuted to life (10 years).
Figure 1. Mean Survival Time to End of Incarceration Period (over legislatively relevant time periods)
Figure 1. Mean Survival Time to End of Incarceration Period (over legislatively relevant time periods)
Figure 1 displays a detailed breakdown of average incarceration lengths over time.
- Murder carried a capital sentence, punishable by death. Releases for sentences commencing during this time period were due to pardons and reduced sentences via the discretionary powers of the Monarch.
- Parole Act results in automatic review of conditional release eligibility for Capital sentences commuted to life.
- Formal split of capital and non-capital murder sentences. Offenders sentenced with non-capital murder are eligible for parole after serving 7 years, while those with capital sentences commuted to life are eligible after 10 years.
- Offenders sentenced with non-capital murder are eligible for parole after serving 10 years.
- Offenders sentenced with non-capital murder are eligible for parole after serving 10 to 20 years (same as those with capital sentences commuted to life).
- Capital sentences abolished and replaced with 1st and 2nd degree murder sentences. First-degree murder is eligible for parole after 25 years and, in some cases, after 15 years. Parole eligibility for second-degree murder sentences vary between 10 and 25 years.
The third and final time period for which incarceration time was examined was from 1976 to 2002. In 1976, the Criminal Law Amendment Act formally abolished the death penalty and introduced the most strict incarceration penalty for murder sentences to date. The mandatory penalty for individuals convicted of what became defined as first and second-degree murder was life imprisonment. Second-degree murder sentences now carry a parole ineligibility period that is 10 years up to a maximum of 25 years, depending on the decision of the sentencing judge. First-degree murder sentences are not eligible for parole before 25 years. However, provisions under the Criminal Law Amendment Act allow for some murder offenders to apply for an early parole review, only after serving 15 years incarcerated.
Figure 1 shows that, historically, the most dramatic effects on the length of incarceration periods for offenders serving sentences for murder occur after 1976. The average incarceration time for 1st-degree murderers was found to be 22.4 years, an increase of 6.6 years over sentences that fell under the capital murder definition. In the cases of 2nd-degree murder, no significant difference in incarceration times for sentences was observed. This may be attributed to cases in which the sentencing judge extended the parole ineligibility to more than 10 years. It is also important to note that survival estimates for this time period are underestimates, as the maximum incarceration time values within this period are right censored.
An examination of the survival curves for each of the three time intervals shows differences in incarceration periods throughout history.
Figure 2. Murder, capital murder and 1st-degree murder
Figure 2 shows that under the general definition of murder, incarceration times were sporadic, ranging from under 9 years to over 40 years. With the capital murder sentence designation and the implementation of the Parole Act, the distribution became less erratic. Implementation of the Criminal Law Amendment Act extended the parole ineligibility period, ‘shifting’ the survival curve outwards, reflecting less releases prior to 15 years. Currently the length of incarceration for the proportion of offenders that will serve over 25 years is unknown.
Figure 3. Murder, non-capital murder, and 2nd-degree murder
Figure 3 demonstrates that, after splitting general murder into two categories, there has been little change in the average incarceration length for non-capital and 2nd degree sentences. Although data for the period from 1976 to 2002 is restricted to the largest event time, and this event time is right censored, we can expect a slight increase in average incarceration time in the future. This would be attributed to 2nd-degree murder sentences receiving parole ineligibility periods closer to 20 years.
Results of this study demonstrate that, in determining the average incarceration time of a murder offender, one must factor in the effects of legislation on sentencing throughout history. Failure to do so may result in errors due to the ranking of observations in inappropriate time periods. This report stratified the population amongst three specific time dependant groups that were based on major legislative changes in sentencing. This facilitated analyses that avoided estimating current censored observations based on historical criteria that no longer apply in sentencing.
Over time, incarceration lengths varied due to differing mechanisms affecting release. Prior to 1961, murder sentences were more likely to end in death or be commuted to life, resulting in a wide-ranging distribution of sentence lengths. From 1961 to 1976 commuted sentences and the new non-capital murder sentences were eligible for parole. The shortest average incarceration times are observed during this time period, particularly for capital cases. From 1976 onward, more stringent criteria have been applied to 1st degree murder sentences, extending the parole ineligibility period to 25 years with the possibility of early parole existing at 15 years for some cases. Average incarceration lengths in this time are the most they have ever been for the more serious murder sentences, while those for 2nd degree murder sentences increased only slightly.
Although the authors went to great lengths to ensure the validity and reliability of the data, information pertaining to inmates who died in prison could be improved upon. Analyses showed minimal differences in survival times while including and excluding cases of death. It was thus assumed that offenders who died did so of natural causes, and were included in the analyses and considered censored. However, a proportion of these deaths could be due to unnatural circumstances; such as murdered while in prison, suicides, drug overdoses etc. These cases may raise average incarceration lengths among the various time periods if adjustments were made to incarceration time based on probability of release. Further, mean survival times were underestimated, particularly for the strata sentenced between 1976 and onward. This is attributed to a lack of observable outcomes for the longest incarceration time in this stratum. The means presented in this paper are thus underestimates of average incarceration times. Prospective research should replicate these analyses with new uncensored observations.
- Date modified :