This paper examines whether the reduction in the percentage of cases where bail is dispensed with (unconditional release) is due to a change in the profile of cases coming before the criminal courts or a general increase in the threshold for dispensing with bail.

A series of statistical (logistic regression) models of the likelihood of unconditional release were constructed to see whether the fall in the likelihood of unconditional release over time remained significant after adjusting for other factors associated with the likelihood of unconditional release. Separate models were constructed for three offences: assault, property crime and property damage.

The report concludes that the reduction in matters where bail is dispensed with is not accounted for by a change in the profile of cases coming before the courts. Instead, this appears to reflect an increase in the threshold for dispensing with bail.