Written evidence from Crest Advisory

Introduction

Long before the COVID-19 outbreak, our justice system was already struggling, as our Swift and Certain criminal justice report (2020) showed. Now a justice system which was already under pressure, is having to adapt to continue to operate at an even more limited capacity. This public emergency is shining a light on the failings of the current criminal justice system and has forced some rapid changes. Crest Advisory was commissioned by the Hadley Trust to explore the impact and legacy of the pandemic on the criminal justice system as a whole. As part of this project, Crest Advisory has modelled the evolution of the court backlog and tested what resources would be needed in order to bring it back down.

The conclusion of our modelling is stark: Covid-19, combined with long-standing legacy issues and the police uplift, will leave the criminal justice system on the brink of a ‘tipping point’, beyond which it ceases to function in a meaningful sense.

This submission explores the following aspects of court capacity:

1. The context: a criminal justice system that was already under pressure
2. Factors affecting the backlog in court modelling: the assumptions[1]
3. The current trajectory: recovering court capacity is not enough to stop the backlog from growing to unmanageable levels
4. To stabilise the Covid-aggravated backlog, the Ministry of Justice would need to double court capacity from a 2019 base lever
5. Other levers that may be pulled to reduce the court backlogs further
6. Long term impacts of a dysfunction in the court system
7. Annex: Modelling methodology

1. The context: a criminal justice system that was already under pressure

The criminal justice system entered the current Covid-19 crisis with a number of long-standing problems. It was underperforming against virtually every measure: police charge rates were going down – decreasing by 47 per cent between 2014 and 2018 from 17 per cent to 9 per cent – against a backdrop of rising crime, court timeliness was getting lengthier with offence to completion timelines having increased by 23 per cent between 2010 and 2018, and reoffending rates were remaining stubbornly high with over a quarter of all offenders reoffending[2] (January 2020). According to our definition, there was estimated backlog[3] in the courts of c.104K cases in March 2019, and prisons and probation were operating at full capacity.

Opportunities to turn around offenders’ lives were routinely missed, victims waited too long for justice and high reoffending rates and ineffective short prison sentences left the public unprotected. Furthermore, serious violent crime has been rising since 2014 and the long term trajectory of police recorded crime is up.

2. Factors affecting the court backlog in the modelling: the assumptions

Our model shows that a variety of factors affect the growing backlog in the courts, including the context outlined above. Factors which are not related to the pandemic, but which have a strong negative impact on the evolution of the backlog include:

• the long-term rise in police recorded crime based on historical trends
• the uplift in police officers which is likely to increase sanction detection levels[4]
• the current time frames for dealing with cases at court that are baked into the system

In addition, the impact of Covid-19 is assumed to be:

• a dip in all 9 indictable and triable either way (apart from fraud and excluding summary offences) police recorded crime offence categories due to confinement recovering to baseline levels in 09/2020
• a 20% spike in violent offences from the 10/2020 to 12/2020 due to anticipated gang/drug distribution power struggles
• a second dip in all 9 offence categories between 11/2020 and 01/2021, modelled as having 50% of the impact of the first wave; levels of recorded crime resume in 02/2020
• a rise in the volume of recorded crime owing to the substantial increase in unemployment from the economic effects of Covid-19; this is modelled to start from 12/2020 when the government support mechanisms (e.g. furlough scheme) come to an end.  This is modelled as 5% rise in violence or sexual offences; 10% for public order/criminal damage offences and 15% for property/theft offences
• a decrease in charging rate by crime type up to -25% due to the CPS Interim Charging Protocol prioritising serious cases and manage pressure on the courts during Covid-19 from 04/2020 to 09/2020[5] (same as the Covid-19 modelled period): - 25%: theft, miscellaneous offences, public order, - 12.5%: robbery, criminal damage and arson, 0%: violence, sexual offences, possession of weapons, drugs.
• court capacity dropping by 90% in March 2020

3. Current trajectory: recovering court capacity is not enough to stop the backlog from growing to unmanageable levels

Based on modelling the main nine categories of indictable and triable either way offences (excluding fraud and summary offences), if court capacity[6] recovered to pre-Covid levels over one year from September 2020, the Crown Court backlog is projected to quadruple by 2024 and the magistrates’ court backlog is projected to  increase tenfold by 2024, compared to the pre-Covid backlogs in March 2020.

Crown Court backlog projections (2014-2024)

if court capacity recovers to 2019 levels over 12 months and if Covid-19’s impact is removed

magistrates’ courts backlog projections (2014-2024)

if court capacity recovers to 2019 levels over 12 months and if Covid-19’s impact is removed

As the graphs above show, court backlogs are exacerbated by Covid-19, but the fundamental dynamics leading to their rise were already in play without the global pandemic. The projected levels of backlog pose a catastrophic risk to public confidence, procedural fairness and effective enforcement of the law.

4. To stabilise the Covid-aggravated backlog, the Ministry of Justice would need to double court capacity from a 2019 base level

Crest’s model explores how court capacity would need to change in order to achieve an equilibrium between the flow of cases coming into the court system (charged and listed) and the court cases exiting the court system (sentenced cases).

It projects that court capacity would need to double, i.e. allowing a throughput of c.35 K sentenced cases (indictable and triable either way and excluding fraud) instead of c.17.5 K sentenced cases per month, in order to stabilise the backlog and bring the flow of cases into equilibrium. The throughput would eventually stabilise in 2024 at c. 29 K sentenced cases per month.

If this is achieved over 12 months from September 2020, the model also calculates the stock of cases in the system and projects that this newly stabilised backlog would be larger than the pre-Covid backlog.

Crown Court backlog projections (2014-2024)

if court capacity recovers to 2019 levels over 12 months, if court capacity doubles over 1 year from 09/20 and if Covid-19’s impact is removed

The Crown Court backlog would stabilise but still be one and a half times larger than the pre-Covid backlog: with a stock of cases waiting to be sentenced in a given month growing from c.45.5K to c.71.5K (x1.5) by 2024.

Crest has modelled a tripling of court capacity, but while this might help stabilise the court backlog faster, it would stabilise in the same place in 2024 as if court capacity was doubled. Thus, tripling resources in the Crown Courts would not be a most effective use of resources (even if it were possible).

magistrates’ courts backlog projections (2014-2024)

if court capacity recovers to 2019 levels over 12 months and if Covid-19’s impact is removed

The magistrates’ court backlog would stabilise but still be three times larger than the pre-Covid backlog: with a stock of cases waiting to be sentenced in a given month growing from c.58.6K to c.208.5K (x3.5) by 2024.

Crest has modelled a tripling of court capacity, which would reduce the magistrates' court backlog further within the model’s timeframe to c.124K cases waiting to be dealt with. This is largely due to the fact that the increase in volume of cases dealt with by the magistrates’ courts will continue to grow at a rate which justifies increased resources, notwithstanding case timeliness.

Crest’s model tests the impacts of increased court capacity, which is the ability to sentence cases during a given month. It does not make assumptions about how this increased capacity is achieved (e.g.  increasing number of physical courts, increasing sitting hours, increasing cases dealt through video-enabled systems etcetera). 

5. Other levers that may be pulled to reduce the court backlogs further

To reduce the Covid-aggravated backlog in Crown Courts and magistrates’ courts beyond what doubling court capacity can achieve, other levers must be pulled to decrease the overall increase in cases and decrease case timeliness.

The increase in the backlog – or volume of cases waiting to be sentenced at any given month – depends on the volume of cases flowing into the system, on the court’s capacity to deal with them and on the courts’ speed in dealing with them (which includes activities carried out by other parties like prosecution and defence staff).

Increasing court capacity can reduce the backlog and stabilise it to a point where inflows match the outflows. However, there will be more cases within the court system at any given month in 2024 than in 2020. Furthermore, average case timeliness has been taken to be consistent, which is a conservative estimate given rising timeliness over the past decade. It is also considered to be stable by this model because it is governed in part by processes and partnerships between criminal justice agencies that are not contained in nationally published data, preventing us from making reasonable assumptions about their evolution.

To bring the backlog back to pre-Covid levels will require a change in more than just capacity. Routes to decreasing the backlog even further include:

• increasing the speed with which cases are dealt with: e.g. increasing the efficiency of listing, decreasing victim attrition, decreasing cracked trials etc.
• decreasing the amount of cases entering the court system by increasing effective out of court disposals
• decreasing the amount of police recorded crime originally entering the CJS through effective crime prevention programmes

These are changes that would require long term policy reform as well as significant investment of resources.

6. Long term impacts of a dysfunction in the court system

The criminal justice system does not function as a unit to the extent that improving performance in one area after it has accumulated a backlog is projected to place pressure on downstream agencies, according to our model. This pressure may be unmanageable given the pre-existing state of the criminal justice system. Thus, decisions made about court capacity should consider the impact on the criminal justice system. Improving the criminal justice system as a whole would require investment and reform across all agencies and aligning missions and targets.

Our model shows that, if court capacity is doubled and equilibrium is achieved in the courts, the prison population will increase by 34% by 2024 which would be a challenge to the current custodial estate (since it is already at full capacity). Crest has calculated that there would be an excess of 15K prisoners by 2024 who would not have a prison place, despite the current 10,000 place increase programme currently underway[7].

Furthermore, assuming equilibrium is achieved in courts, suspended sentence orders are projected to increase by 24%, post-release supervision caseload will increase by 30% and community sentence orders are projected to increase by 14% by 2024. This will put extreme pressure on the probation which was already underperforming and struggling to such an extent that a re-nationalisation was planned for December 2020[8].

7. Annex modelling methodology

In order to understand the scale of the impact and of the necessary resources to deal with the crisis, Crest has developed a stock and flow model of the criminal justice system:

Design and aim:

●            Crest have developed a stock and flow model that looks at the CJS as a whole focusing on nine major categories of offences, excluding summary offences and fraud, over the time period of 2014 to 2024

●            The model tracks how major offence cases flow through the CJS and the impact on the ‘stock’, projecting the impact on capacity and/ or outcomes ‘flow’.

●            The purpose of the model is to support a conversation about the scale of the impact of Covid-19 crisis on an already struggling CJS and about what investment and reform is required to maintain a functioning system

Sources:

•            Police recorded crime Police Force Area open data tables, year ending March 2013 onwards. Updated: 23rd April 2020; Next scheduled update: July 2020. www.gov.uk/government/statistics/police-recorded-crime-open-data-tables

•            Crime Outcomes in England and Wales Open Data, April 2019 to December 2019 and previous publications. Updated: 23rd April 2020; Next scheduled update: July 2020. www.gov.uk/government/statistics/police-recorded-crime-open-data-tables

•            Police local data released monthly:  https://data.police.uk/data/ . Covers street-level crime, outcome, and stop and search data in CSV format, containing detailed crime data and information about individual police forces and neighbourhood teams. All the data on this site are made available under the Open Government Licence v3.0.

•            Criminal justice statistics covering the operation of the magistrates’ and crown courts, including information on prosecutions, convictions and sentencing decisions. Versions published 21 May 2020 and last updated 28 May 2020. Unfortunately, data are not published in a consistent way between editions.  Fuller datasets cover more detail and published in the December 2018 and December 2019 editions of these statistics at:  https://www.gov.uk/government/statistics/criminal-justice-system-statistics-quarterly-december-2019

•            For prisons and probation, we used offender management statistics, which covers flows into these services (receptions into prison or probation starts) and flows out (discharges from prison or probation terminations) as well as the caseload of both services at specific points in time. https://www.gov.uk/government/collections/offender-management-statistics-quarterly

Limitations and validity:

●            Any model is reliant on the accuracy of the data that underpins it - this model is no different. It relies on publicly available data which does not cover the past 6 months and does not detail court processes beyond milestones in the case journey (listing, first hearing, sentenced) and excludes individual case timeliness’ variables like listing priorities, victim and witness attrition, defence and prosecution communication etc.

●            Crest has been transparent on the model’s assumptions. They have been made based on data and expert judgement. Real data may vary if events differ from the assumptions that were included

●            A model’s usefulness is determined by how well the underlying dynamics is captured and is representative of the drivers, capacities and other constraints. This model is obviously a simplified version of reality, but the tests of projected numbers are coherent with recorded ones

October 2020

11