FORUM on Corrections Research

Examining the Unexamined: Recidivism Among Female Offenders

This article makes a first attempt at examining recidivism among federal female offenders. Female offenders make up a much smaller proportion of the inmate population than male offenders. Because of this, research on female offenders is not as common as that on male offenders. In this article, we hope to make a small contribution toward understanding recidivism among female offenders.

In Canada, persons sentenced to terms of two years or more serve their sentence in a federal institution; sentences of less than two years are served in provincial institutions. In this study, we included all women serving their first sentence in a federal institution who were released in the 10-year period between 1 January 1978 and 31 December 1988. We followed up these 968 offenders until 30 June 1993 to find out whether they were readmitted to federal custody at any time following their release.

Of the 968 released female offenders, 213 were returned to federal custody during the follow-up period for an overall recidivism rate of about 22%. The Sample Most (81.4%) of the 968 female offenders in the sample were non-native; 13.7% were native. Information was missing or not known for 4.9% of the sample.

When first admitted to federal institutions, these offenders were between 17 and 71 years old; the average age was 30. About 40% of the offenders were between the ages of 18 and 25, another quarter (25%) were between 26 and 30, and about 28% were between 31 and 45.
The most common major admitting offences (the one for which they received the longest sentence) were drug-related (29.1%), homicide-related (17.7%), robbery (16.5%) and property-related (11.9%). (Please note that homicide-related offences include manslaughter, first- and second-degree murder, and capital and non-capital murder.) Only 1.2% were sex offences.

About two thirds (66%) of these offenders were serving sentences of two to five years. Another fifth (20%) were serving from more than 5 to up to 10 years. About 3.7% were serving sentences of more than 10 years, and 8% were serving life sentences. Release Almost two thirds (64.3%) of the sample were released on full parole and about one quarter (23.9%) on mandatory supervision (now called statutory release). The rest were released at the end of their sentence (0.7%) or by other means (8.3%), such as transfer to a provincial institution.

Generally, an offender who is considered a safe release risk will be released on full parole. If not, the correctional authorities and the Parole Board will stipulate that the offender remain in the institution until the date arrives for release on mandatory supervision.

In this study, we found non-native offenders were more likely than native offenders to be released on full parole (about two thirds versus only about half). Conversely, non-native offenders were less likely than native offenders to be released on mandatory supervision (22.7% versus 37.2%). Readmission or Recidivism Our definition of recidivism was broad: anyone who was released into the community and subsequently readmitted to a federal institution. This includes, for example, not only those who were readmitted for new offences but also those who were readmitted for technical violations of their release conditions (with no new offences). Please keep in mind the broad nature of this definition when considering the following results.

Of the 968 female offenders released into the community, 213 (or 22%) were readmitted. Figure 1 shows the percentage of offenders who never returned to federal custody (as of 30 June 1993) and the percentage of those who were readmitted for various reasons. As shown, 78% of the women in the sample never returned to federal custody. In addition, 5.6% were only readmitted after their sentence had ended and they were no longer under correctional supervision. This bodes well for our ability to manage female offenders.

Of the 213 women who were readmitted to federal custody during the time period under study, half (49.8%) were readmitted for technical violations of their release conditions⁽²⁾ (i.e., release revoked without a new offence). Another quarter (25.4%) were readmitted with a new warrant of committal; that is, their previous sentence had ended and they were no longer on conditional release, then they committed a new offence and received a new federal sentence. About one fifth (21.1%) had their release revoked for a new offence. Lastly, about 3.8% were readmitted for other reasons.

---

Figure 1
Race and Recidivism Native female offenders in this study were over represented in the group of women who were readmitted. While native women made up only about 14% of the study sample, they made up about 27% of the recidivists. And while non-native women comprised about 81% of the sample, they comprised 71% of the recidivists.

Put another way, about 44% of native female offenders in the sample were readmitted, compared with about 19% of non-native female offenders. Age and Recidivism We found that younger offenders were more likely to be readmitted than older ones. We calculated the offenders' age at the time of release and found that as the age category increased, the percentage of recidivists decreased. For example, 29% of those between the ages of 18 and 25 were readmitted, compared with 22% of those between 26 and 30, 20% of those between 31 and 45, 16% of those between 46 and 60, and 11% of those over 60.

Those who were readmitted were, on average, 32 years old when they were readmitted. Original Admitting Offence and Recidivism What types of female offenders are recidivists? Our data base allowed us to look at the original major admitting offence for the female offenders who were readmitted.

Of the 213 women who were readmitted, about 22.5% had originally been incarcerated for homicide-related offences (manslaughter, first- and second-degree murder, and capital and non-capital murder). Another 22.5% had been incarcerated for robbery, about 16% for property-related offences, 8% for drug-related offences and the rest for other offences.

Comparing these results with the major admitting offences for the original sample, we found that women with drug-related offences were under represented in the sample of readmitted women, while those with robbery and property-related offences were over represented.

Women with homicide-related offences were also over represented. Although they made up only 17.7% of the original release sample, they accounted for 22.5% of those readmitted.

We thought this latter finding was interesting, so we investigated the homicide-related offence group further. We found that while the murder offenders in this group were actually under represented in the sample of readmitted women, the same was not true for manslaughter offenders. Although they accounted for only 13.5% of the original sample, manslaughter offenders accounted for more than one fifth (21.1%) of those readmitted. In fact, just over one third (34.3%) of the released manslaughter offenders were readmitted to federal custody. This was the highest read-mission rate for any offence group. Release Type and Recidivism We also examined whether the type of release was related to subsequent readmission to a federal institution. As expected, we found that offenders released on full parole fared much better than those released on mandatory supervision. Offenders released on full parole made up almost two thirds (64.3%) of the total sample, but they comprised less than half (46.5%) of those readmitted. On the other hand, offenders released on mandatory supervision comprised about 24% of the sample, but they made up about 44% of recidivists.

Looked at another way, of all the offenders released on full parole (622), only about 16% were readmitted. This compares with about 41% of the 231 offenders released on mandatory supervision.

We can also examine the prognosis for offenders on different types of release by studying the reasons for their readmission to federal custody. We looked at readmissions for technical violations of release conditions, new warrants of committal and revocations for a new offence. Figure 2 shows the results of this analysis.

Overall, recidivists who were released on full parole or mandatory supervision were most likely to be readmitted for a technical violation rather than for a new offence (a new warrant of committal) or for a release revocation because of a new offence.

If you look at Figure 2 closely, you will see that recidivists released on mandatory supervision were more likely than those released on full parole to be readmitted with a new warrant of committal (24.5% versus 14.1%) but less likely to have their release revoked with a new offence (17% versus 29.3%).

This finding makes sense because release on mandatory supervision occurs later in an offender's sentence than release on full parole. An offender who has been released is only on "conditional" release (and under the supervision of correctional authorities) until the sentence has ended. After that, we have no jurisdiction to revoke release.

Given this, if an individual who was originally released on mandatory supervision committed a new offence, it is more likely (than for an offender originally released on full parole) that the original sentence will have ended by the time the new offence is committed, and thus, the offender would be readmitted on a new warrant of committal rather than for a release revocation with a new offence.

We found some interesting results when we looked at recidivists in the "other" release category. This group included offenders released from federal custody because they were transferred to another jurisdiction or because they were freed by a court order. In fact, 18 of the 20 recidivists in the other category fell into the latter group. The vast majority (85%) of recidivists in the other group came back with new warrants of committal.

---

Figure 2
When Do They Come Back? We calculated how long the recidivists had been released before being readmitted. Figure 3 shows the break-down by time periods.

Almost two thirds (63.6%) were readmitted within two years of release, but about 14% were readmitted more than five years after release. While these findings underscore the need for close monitoring of offenders during the first couple of years after release, it also suggests that long follow-up studies are necessary to track recidivism among female offenders.

We also looked at how many women were readmitted after their original sentence had ended (i.e., after the warrant expiry date of their first sentence). Of the 193 women on whom information was available,⁽³⁾ 83 (43%) were readmitted after their warrant expiry date.

At first, this figure of 43% appears to be incorrect since only 25.4% of the sample were readmitted with a new warrant of committal and most of the rest were readmitted because their release had been revoked.

This apparent discrepancy can be explained. If an offender commits an offence while on conditional release, their release will be revoked with a new offence. But, sometimes these offenders will spend time in a local jail waiting for their new offence to be heard, so by the time they finally enter the doors of the federal institution, their original sentence has ended (because the clock keeps ticking regardless of where they spend their time). Even though these women don't enter a federal institution until after their original sentence ends, they are still counted as having had their release revoked with a new offence.

---

Figure 3 What Types of New Offences Do They Commit? We studied those who were readmitted with new offences - that is, those who had their release revoked with a new offence or were admitted with a new warrant of committal - to see what types of offences they committed while on release.

Of the 99 women in this subsample of recidivists, about one quarter (24.2%) committed property-related offences, almost one fifth committed homicide-related offences⁽⁴⁾ (18.2%) or robbery (17.2%), 7.1% committed drug-related offences and the rest committed other offences. Multiple Recidivism We also very briefly looked at how many of the released offenders were readmitted more than once. That is, how many were released then readmitted, released then readmitted, etc., during the time period under study?

Again, of the 968 female offenders released, 22% (or 213) were readmitted at least once. In fact, 14.5% (140) were readmitted only once, 5.2% ⁽⁵⁰⁾ were readmitted twice, 1.1% ⁽¹¹⁾ were readmitted three times and 1.2% ⁽¹²⁾ were readmitted four times or more.

We can look at this another way. If an offender has been readmitted once and then been released again, what are the chances of further recidivism? Of the 213 women who were readmitted at least once and then released, about one third (34.2% or 73) came back. Of these 73 who were readmitted twice and then again released, almost one third (31.5% or 23) came back yet again. And of these 23 women who were readmitted three times and then released, about half (52.1% or 12) came back at least one more time.

Looking only at those who were readmitted more than once (73 offenders), native offenders were again over represented, comprising 28.8% of the multiple recidivists but only about 14% of the whole sample.

We also found that offenders whose original offence was robbery were over represented in the group of multiple recidivists and drug offenders were highly under represented. Although robbery offenders comprised 26% of multiple recidivists, they only comprised 16.5% of the original sample. Conversely, drug offenders comprised only 8.2% of multiple recidivists but made up 29.1% of the original sample. Summary Overall, the readmission rate for federal female offenders released between 1978 and 1988 was 22%. About half of these readmissions were for technical violations of release conditions and most occurred within two years of release.

Female offenders who were native, younger, on mandatory supervision or had committed robbery or manslaughter were over represented in the readmission group. Those who were non-native, older, on full parole or had committed drug-related offences were under represented.

Offenders who committed offences while on release were most likely to commit property-related offences and were less likely to commit drug-related offences.

---

⁽¹⁾The authors of this article may be reached at the following address: Research and Statistics Branch, Correctional Service of Canada, 340 Laurier Avenue West, Ottawa, Ontario KJA 0P9.
⁽²⁾Some offenders in this category could have been readmitted for a technical violation of their release conditions and then, after they were already readmitted, been convicted for an offence they had committed while on release. These offenders would still be grouped with offenders readmitted for a technical violation, even though they more accurately belong with offenders readmitted for a new offence.
⁽³⁾Information was missing on 20 cases.
⁽⁴⁾Most (13 of 18) of these homicide-related offences were manslaughter.