STATISTICAL GAPS LEAVE ROOM FOR SPECULATION

The Ministry of Justice has been accused of many errors in policy and practice but what it cannot be accused of is the production of statistics.  We might not be told the ethnicity of the magistrates on any one sitting but we do have information of the magistracy as a whole.  We do not have or need to know the age or sex of those advocates appearing for a defendant but myriad gigabytes  are in the ether telling us of eg  the elapsed time from a defendant`s first appearance to sentence or indeed the delay from charge to that first appearance.  A major court function where there appears to be a gap in published information is the failure-to-appear (FTA) rate among defendants granted bail.

The MOJ isn`t short of employees being paid to provide the numbers that politicians and others pour over relentlessly.  The MoJ's Data & Analysis Directorate is described as a multi-disciplinary team of around 500 staff, covering analysts, statisticians, researchers, economists and data scientists across the department. A precise breakdown of how many of those 500 are specifically classified as statisticians (as opposed to other analytical professions) doesn't appear to be publicly published in an easily accessible form. The closest public data would be the Government Statistical Service (GSS) workforce statistics, which track statisticians across all government departments.  Nevertheless we do know that the approximate numbers granted bail at magistrates' courts were:

The question of how many failed to answer bail and had warrants issued is another matter. The Ministry of Justice does not publish a direct national statistic linking magistrates court bail grants to subsequent failures to appear. However, a recent investigation reported that nearly 60,000 arrest warrants for defendants failing to attend court were issued in England and Wales in the most recent year, with the annual figure having risen by about 50% since 2020. Although the Ministry of Justice no longer appears to publish a current annual figure there are some useful historical benchmarks. A Parliamentary Answer stated that across all courts in England and Wales, approximately 11–12% of bailed defendants failed to appear during the late 1990s and early 2000s. For 2006 the figure was about 10%. Ministry of Justice statistics noted that FTA warrants had been falling and that by 2012/13 police forces received about 70,400 FTA warrants annually with 88% executed. The MoJ expressly linked this to declining numbers of defendants who failed to appear on bail or summons. Trying to make some sort of sense of the available numbers is that a single defendant can fail to attend more than once and not every failure to appear arises from a defendant who was granted bail. Around half a million defendants were granted bail in magistrates courts over 2022–2024. Historical Ministry of Justice and NAO evidence suggests that somewhere between one in ten and one in seven bailed defendants may fail to attend at least one hearing, implying a broad range of roughly 50,000–75,000 failures to appear over a three-year period, but no current official national statistic appears to be published.

A Parliamentary Answer in 2021 disclosed HMCTS management information on FTA warrants. The answer confirms that HMCTS can extract warrant data directly from court systems and that the figures are management information rather than official statistics. Some police forces have recently released data on outstanding FTA warrants.  Dyfed-Powys Police reported 108 outstanding FTA warrants as of January 2026 of which 95 related to magistrates court matters and 13 to crown court matters. West Yorkshire Police reported 2,148 outstanding FTA warrants as of January 2026 covering Bail Act and magistrates  court warrant categories.

The Ministry of Justice publishes numbers granted bail,  numbers remanded in custody and the numbers of FTA warrants. It does not appear to publish a dataset linking those events at defendant level. Consequently, there is no routinely published national figure saying: "X% of defendants granted bail at magistrates courts failed to surrender and had a warrant issued."  The key historical figures published by HMCTS were:-

Currently there is no simple straightforward answer to the question:-  of those granted bail how many failed to answer that  bail?

To sum up; Roughly half a million defendants were granted bail by magistrates' courts during 2022–2024. Magistrates courts have historically issued tens of thousands of FTA warrants annually, typically around 90,000 in the published HMCTS series. The Ministry of Justice does not publish a defendant-level statistic showing what percentage of bailed defendants subsequently fail to surrender.


With all the facts at its disposal yet such a gap in the published statistics one wonders why such basic information is not available for all to see and comment upon. Where there is such a deficit there is always room for speculation by those with an interest in the subject.

IT`S ALL GREEK TO ME

I am soon travelling to warm pastures where sun, sea and sand will be my daily view for breakfast al fresco and antikristo will be accompanied by a bottle of liatiko at dinner.

Accordingly tomorrow`s post will be the last until early July.

 τα λέμε τότε

ENGLAND 12 SCOTLAND 15 BUT STILL NOT PROVEN

The title of today`s post does not reflect the result of an international rugby match: it is the number of jurors in criminal trials in England and Scotland. I have long favoured the long standing practice of Scottish criminal courts in having at the jury`s disposal the "3rd verdict" namely not proven that was available until Victims, Witnesses and Justice Reform (Scotland) Act 2025 received Royal Assent on 30 October 2025 and is in force with staged commencement. To appreciate this fundamental change in Scotland in comparison to the current decision making system in England some historical facts are helpful.

The English criminal jury of 12 emerged gradually during the 12th–14th centuries as part of the common law system. By the late Middle Ages 12 had become the settled and accepted number. It was not originally fixed by a single statute; rather it became entrenched through common-law practice. The number was regarded as so fundamental that courts later treated it as an essential characteristic of a jury.

Scottish criminal juries settled at 15 members by the late 16th century. Unlike England, Scotland developed its own legal tradition and never adopted the English 12 person criminal jury. The number 15 became a distinctive feature of Scots criminal procedure.  

To understand how the neighbouring countries developed such a fundamental difference  Scotland retained its own distinct legal system after the Acts of Union 1707 and its larger jury number likely reflects older continental European civil law influences on Scots procedure and perhaps deliberate retention of distinctiveness from English practice post Union.  In essence the differences were well established before the two kingdoms were politically united. 

Majority verdicts were introduced by the Criminal Justice Act 1967 before which a criminal jury in England and Wales had to be unanimous.  This change allowed a 10:2 verdict after the jury has deliberated for at least two hours.  The main reasons for such a change were the fear of jury nobbling or corruption, intimidation and the ability of a single juror to force a retrial.

For many people this previously ethereal legal matter  has become a social and political point of argument. Of equal importance one might inquire if there is a statistical difference in mistrials and hung juries between Scots and English juries.  The recent second mistrial of airport attackers has brought this previously little regarded aspect of the English jury system under the spotlight as has the recent spate of perverse verdicts. 

In the English procedure even after the 1967 reforms  juries can still fail to reach a verdict eg 9:3, 8:4, 7:5.  Thus England necessarily experiences more mistrials arising from jury deadlock than Scotland.  Under the traditional Scottish system 8 votes for guilty equals conviction and anything less is an acquittal. Scotland's three verdict system historically allowed jurors uncomfortable with outright acquittal to reach for not proven which arguably reduced hung juries by providing an escape valve.  Below is an interesting comparison. The conviction thresholds are surprisingly different.  

The abolition of not proven was one of the most controversial Scottish criminal justice reforms in decades and there was significant opposition from parts of the legal profession.  Not proven and not guilty had exactly the same legal effect: the accused walked free from court.  Critics argued that if two verdicts have identical legal consequences but very different social meanings the system lacks clarity.  Victims' organisations argued that the verdict denied closure because it seemed to acknowledge suspicion without conviction.  Defence lawyers pointed out the opposite problem: acquitted defendants could remain permanently stigmatised because the verdict appeared to imply lingering guilt.  Campaign groups argued that not proven was used disproportionately in rape and sexual assault trials and may have contributed to lower conviction rates.  Thus under pressure Scotland moved from 8 of 15 (53.3%) to convict to 10 of 15 (66.7%) to convict.

Many[a majority?] in the legal profession objected to the abolition of the third verdict.  They argued in essence that the reality of a juror`s thinking was now subsumed in translating that to a convenient legal process.  However the change was effected by intense lobbying by victims' groups particularly by those concerned with sexual offences. The higher 10:15 conviction threshold reassured many MSPs that defendants were not losing protection without receiving something in return.

The results of these changes will be analysed in the years to come. Considering a legal system must attempt to deter, punish and rehabilitate those who break the boundaries set by government the benefits of this profound change in a 300+ year old tradition must for some time to come be considered not proven.