International Journal on Criminology Volume 1, Number 1, Fall 2013 | Page 42

Demographic Analysis of the Penal System
CROSS-SECTIONAL ANALYSIS: cross-sectional analysis involves phenomena that
appear during a set period of time, usually a calendar year, within a set of cohorts. 9 It is
fundamentally different from longitudinal analysis. Here is an example: consider a cohort
of persons committed to prison. Through observation over time, we can learn who among
the inmates of this cohort will benefit from temporary leave during their detention. This
follow-up can persist up until their release. By ranking these leaves (1st leave, 2nd, etc.),
we can analyze the conditions under which these leaves took place (without incident, late
return to the detention facility, escape). This would be a longitudinal analysis.
However, in a cross-sectional study, we might also look at all the leaves granted in a
particular year to the population of convicted detainees. These detainees would belong to
an entirely different set of imprisoned cohorts and would represent a heterogeneous
population with regard to the length of time already spent in prison.
OBSERVATION OVER TIME: Unlike retrospective cohort analysis, which looks to
the past, the observation of a cohort over time looks toward the future. As an example, let
us look at one of the first correctional studies carried out in France using this method. It
involved the 6,745 entries into detention in February 1983. 10 In cases where one person
entered several times, only the first entry of the month was considered. Thus, we can in
fact refer to them as incoming inmates. The study began in 1983 and involved three
periods.
During the first period, we analyzed the socio-demographic and criminal structure of
this cohort. After determining a representative sample of these incoming inmates
(n=1326), we allowed for a sufficient period of time to pass so that most incoming
inmates were released. Following a few tests, we decided to end this period of
observation after 27 months: 95% of the incoming inmates had been released at that
point. Fifty percent had been released after 2.5 months (median of the duration of
detention). It was then possible to analyze the time spent in prison: the schedule of exits,
the study of detention, of pre-trial detention and of (short) sentence modification.
The third step of the observation over time involved examining the individual criminal
records of the sample, on average five years after release. For those released before a
ruling was issued, it was possible to determine whether they had ultimately been
convicted and if so, of the nature of the sentence. For all cases, we could determine
whether they had been involved in a new offense after their release, and whether they
were convicted (study of the cohort's judicial follow-up). Fifty-nine percent of those
released were involved in a new case, punishable by a penalty of some kind, and the rate
of return to prison was 39%.
5. Prison population inflation and overpopulation of prisons
PRISON POPULATION INFLATION: Any mention of prison population inflation
suggests that the increase in the number of persons in prison is “very high” (stock data).
In other words, it does not correspond with the increase in the number of inhabitants. 11
!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
9 Tournier, “Note technique.”
10 Barré, Tournier, and Leconte, La mesure.
11 Council of Europe. Prison overcrowding and prison population inflation, recommendation No. R
(99) 22, adopted by the Committee of Ministers on September 30, 1999, and report prepared with
41