Scottish Prison Population Projections: February 2026
This report presents short-term Scottish prison population projections for the six month period from February 2026 to July 2026.
Terminology definitions
This section includes definitions and explanations of modelling practices and terminology used to describe the level and range of projections for the total, sentenced and remand prison populations.
Microsimulation scenario definitions
As in the first publication of June 2023, three scenario variants are reported: 'low', 'central', and 'high' court throughput. It is assumed that remand arrivals will remain similar to the levels observed in the twelve months leading up to February 2026.
All scenarios use common underpinning data on court case progression and the prison population, sourced from the SCTS and the Scottish Prison Service, respectively. The scenarios differ only in their assumptions about how future levels of court case conclusion rates may evolve.
The microsimulation methodology requires each scenario to be simulated repeatedly – dozens of times – to generate prediction intervals for the total, sentenced and remand populations. The full projection includes the 95% prediction intervals for all three scenarios. These intervals are shown in Figures 1, 14, 15 and 18, and are overlapping, such that it is not possible to easily identify individual scenarios visually.
The upper and lower estimates of the populations in Table 2 represent the outer limits of the 95% prediction intervals for the most divergent scenarios.
Court throughput rate scenario definitions
The 'central' scenario assumes that the average case throughput per court will remain similar to the average observed over the year leading up to 1 February 2026 data cut-off.
The 'high' scenario assumes that the average case throughput per court will be 10% greater than observed.
The 'low' scenario assumes that the average case throughput per court will be 10% lower.
A rolling sampling period is used, meaning each edition of the projections incorporates case conclusion data from the most recent twelve-month period. This ensures that the model reflects current trends in each successive update.