Scottish Prison Population Projections: February 2026

This section provides an overview of the assumptions and scenarios used to project the prison population, and explains how the model’s previous projections are validated against actual prison population figures to assess robustness. The first issue of the Scottish prison population projections[1] provides a more detailed overview of the modelling technique used[2].

Model Assumptions and Scenarios

The projection scenarios are based on a range of assumptions about how the rate of transitions into and out of the prison population might change. The assumptions reflect trends and planned changes in the court system, including: changes in court capacity; variations in court conclusions rates; fluctuations in remand arrivals; the remand/bail mix; and the prioritisation of remand case progression through courts.

Departures from the sentenced population are estimated using a combination of court disposal data and snapshots of the prison population.

To assess the sensitivity of the prison population to variation in court case conclusion rates, three scenario variants are included in the modelling:

• 'Central' throughput: based on case conclusion rates per courtroom from February 2025 to January 2026;
• 'Higher' throughput: assumes a 10% increase in the average case throughput per court; and
• 'Lower' throughput: assumes a 10% decrease in the average case throughput per court.

The three scenarios are summarised in Table 1.

Table 1. Prison population scenario variants.

The COVID-19 pandemic disrupted the justice system, with criminal court business temporarily restricted to essential business only[3]. Combined with increased court activity, this led to a backlog of trials. While court recovery resources have reduced the backlog from its peak, they remain high in High Court and Sheriff Solemn compared to pre-pandemic levels. This imbalance in the justice system makes it more difficult to estimate future case conclusion rates and imprisonment disposals.

Therefore, since the progression of the justice system’s recovery (e.g., rate of reduction of scheduled trials) impacts the model’s assumptions and longer-term predictive power, the projections cover a limited period – from February 2026 to July 2026. Assumptions for these months are sampled from the twelve-month period from February 2025 to January 2026.

Model Quality Assurance

At each update, the previous projections are compared with the actual prison population figures to determine the model’s continued suitability to produce reliable projections. Figures 14 and 15 show the projections from the microsimulation which were published in July 2025[4] and October 2025[5].

Figure 14. Prison population projections and the actual prison population beginning on 1 July 2025, based on courts and prison population data up to the end of June 2025. The actual population is shown as a series of black points up to 31 December 2025.

Even after accounting for a wide variety of uncertain dynamics in the system, Figure 14 indicates that the projected ranges for the remand, sentenced and total populations have aligned reasonably well with the observed population levels. The three populations were tracking the ‘low’ scenario throughout the projection, until mid-November 2025 when EER measures were initiated. These projections did not include the impact of EER as the measures were not yet agreed. Notwithstanding this, the actual and simulated populations remained well aligned throughout the projection.

Figure 15. Prison population projections and actual prison population beginning on 1 October 2025, based on courts and prison population data up to the end of September 2025. The actual population is shown as a series of black points up to 23 February 2026.

Figure 15 shows that the prison populations have also aligned with the latest projections published in October[6]. Throughout the projections the total population has remained within the prediction intervals. The actual sentenced population tracked the lower estimate in October and November, but aligned with the central conclusions scenario estimate from New Year onwards. The remand and total populations have largely tracked the lower conclusions estimate.

To further assess the model, back-casting is used. This involves generating a retrospective projection using actual monthly court throughput data, thereby removing most assumption-based uncertainty. Any divergence between the back-cast and the actual figures may indicate technical limitations in the model. In this publication we have provided two back-casts – one including the initial EER period and one covering a six-month period where no emergency releases took place.

The most recent back-cast, presented in Figure 16, indicates that the actual prison population remained within the modelled range from August 2025 to January 2026, with the total population slightly higher than projected.

Figure 16. The back-cast is based on remand arrival and court throughput data from August 2025 to January 2026. The actual population is shown as a series of black points.

An earlier back-cast, prior to the emergency early release tranches, shows strong agreement for all three populations with the actual populations.

Figure 17. The back-cast is based on remand arrival and court throughput data from May 2025 to October 2025. The actual population is shown as a series of black points.

Microsimulation Model Limitations

The model relies on access to a large amount of high-quality, frequently updated data on court activity and prison populations. Some of this data can be resource intensive to obtain and process.

The model does not currently simulate flows by crime type, so crime-specific trends are not explicitly modelled.