These statistics are categorised as Under Reservation. This categorisation indicates that the quality of these statistics do not meet the standards required of official statistics published by the CSO.
For further information please refer to the Under Reservation FAQ page.
1. What do Prison Re-offending Statistics measure?
Re-offending statistics measure the level of recorded re-offending by offenders who received a custodial sentence following their release from prison. The three year re-offending rate indicates the percentage of people who were convicted for a crime incident that was recorded within three years of their release. A further two years after the three year period is set as the limit for a valid conviction to take place. The one year re-offending rate indicates the same measure but includes only those who were convicted for a crime within one year of their release, with a further year set as the time limit for the court outcome to take place for example:
2. Are these statistics "Under Reservation"?
These statistics are produced using a combination of An Garda Síochána and Irish Prison Service records. All statistics produced by the CSO that utilise the PULSE dataset are classified as Under Reservation. The classification has been applied to reflect the concerns of the CSO as to the completeness and accuracy of PULSE data. Further information on the rationale for, and the implications of, the Under Reservation classification may be found here:
3. How is the Prison Re-offending rate calculated?
In order to calculate the rate of re-offending, it is necessary to match records from data collated by the Irish Prison Service with the Garda PULSE data. The matching process results in the selection of a PULSE identifier (or identifiers) for each subject. Convictions data is then examined to determine if individuals, who served either a prison sentence or received a fine sentence, had any subsequent convictions. In addition to these PULSE indicators, if an individual indicates from prisons data that a re-committal to prison takes place within the qualifying period then this also is used as an indication of a re-offence.
4. How are records matched?
Due to the absence of a unique identifier in the Irish Criminal Justice system, the matching process involves the comparison of individual records. In the region of 90% of all matches are made automatically. Due to data quality issues and limitations on the reliability of some information provided to law enforcement authorities, the remaining 10% of matches require manual examination. Methodological improvements to the matching process were introduced for the 2011 to 2018 cohorts. These improvements involve the introduction of new software and an updated matching protocol. The improvements have increased the reliability of the matching process but, in the absence of a common identifier in the criminal justice system, there will always be a degree of subjectivity associated with the matching process.
5. Are the rates of Probation and Prison Re-offending comparable?
The Probation and Prison publications were developed separately. They were designed primarily to meet the information needs of the Irish Probation Service and the Irish Prison Service. The needs of these organisations are broadly similar but there are some differences. For example, convictions must be secured within 24 months for three year re-offending and 12 months for one year re-offending of the date that the crime incident was reported to be included in the calculation of the prison re-offending rate. No such requirement exists with respect to the calculation of the probation re-offending rate. Certain offences such as traffic offences are not included in the calculation of the rate. The exact list of exclusions differs somewhat between the publications. It is also important to bear in mind that the characteristics of the cohorts under consideration differs. The prison cohort tend to have committed more serious offences than their probation or fine sentence counterparts. Due to the differences discussed, it is not possible to make direct comparisons.
6. Is it possible to make comparisons across jurisdictions?
The definitions used in re-offending studies are not standardised internationally. The re-offending window, that is the time period after release from prison during which re-offences are considered, varies across jurisdictions. There are also variations as to what constitutes a re-offence. In some countries a custodial sentence is necessary while other countries consider rearrest to be the most appropriate measure. It is extremely difficult to compare re-offending rates across jurisdictions due to differences in definitions, data quality, legal systems, legal culture and justice policy. Controlling for these differences would be a complex, but essential, process to draw meaningful conclusions from comparisons.
7. What's the difference between prison and fine sentence re-offending?
Prison re-offending measures the identified level of re-offending that takes place once a person has been released from a custodial sentence. Fine sentence re-offending measures the identified re-offending rate for the mostly non-custodial sanctions called fine sentences. A person who receives a fine sentence does not for the most part serve a custodial period. A very small proportion of fine sentences (less than 10%) result in a custodial period and the periods of detention if they do occur are usually very small (less than five days).
8. What's the difference between three year and one year measures of re-offending?
Three year re-offending allows for a three year period after a prison sentence or fine sanction comes to an end to monitor for any re-offences. If an incident is identified a subsequent two year period is reserved to see if the incident receives a qualifying court outcome. To measure one year re-offending, just a one year time period for the re-offence and also one year for a court outcome is used to measure qualifying incidents.
Learn about our data and confidentiality safeguards, and the steps we take to produce statistics that can be trusted by all.