Arrests for notifable offences under PACE 1984
The Home Office’s Police powers and procedures England and Wales statistics break down arrests by local police force and age category. Police forces already have to collect and provide this information to the Home Office, so no extra work is involved. This makes it relatively simple to assess trends in arrests over time.
The statistic are normally published in October for the year up to the preceding March. For example, the arrest statistics up to March 2017 are available here.
The data is provided at a national level and broken down (e.g. by force, sex and age)
Non-notifiable offences under PACE 1984
A limitation of the Home Office data is that includes only offences that are considered serious enough to be ‘notifiable’ offences. Non-notifiable offences do not need to be recorded for statistical purposes. These include minor motoring offences, TV license evasion and drunk and disorderly.
This is important for deveoping an AA scheme because non-notifiable offences are not insignificant in terms of numbers, as indicated by the table below.
Year ending December 2016
Penalty Notice for Disorder
Arrests for terrorism-related offences
In addition to criminal matters dealt wiht under the POlice and Criminal Evidence Act 1984 (PACE), a small number of people are arrested under specific terrorism legislation. This makes up a very small proportion of arrests and detentions. For example, the total number across England and Wales in the year to September 2017 was 400 (of which 18 were aged under 18).
Despite the small number of cases, it is important to include provision for terrorism related cases. They are often (though not always) very serious cases and can require longer periods of support.
The Home Office regularly publishes statistics on arrests under the Terrorism Act 2000.
Refusal to authorise detention
It should also be noted that arrest figures are not exactly the same as custody detention figures. This is because it is possible for a police custody sergeant to refuse to detain a person who has been arrested and brought to their custody suite. However, the figures are likely to be very similar.
Local police forces should be able to provide more recent and accurate data on arrests and detentions. If possible, this should include consideration of non-notifiable offences and the extent to which they involve detentions or voluntary interviews.
- Home Office: latest arrest statistics
- Office for National Statistics: latest crime and justice statistics
- Ministry of Justice: criminal justice statistics quarterly
 Office for National Statistics (2017), Crime in England and Wales: Bulletin Tables year ending March 2017
Reliable data on vulnerable adults may be challenging to locate and access. Home Office statistics do not provide data on the number of adults detained or interviewed who either should have been or was supported by an AA .
Police forces may be able to provide an insight into levels of demand. However, the quality of this data will be limited for three main reasons:
- The challenge faced by police in identifying mental vulnerability; and
- AAs are not always called even where mental vulnerability is identified;
- The capacity of police information systems.
Academic studies suggests that mental vulnerability is under-identified in police custody. Figures provided by police forces across England and Wales suggest that an average 3% of detentions of adults involve an AA.However, this ranges from below 1% to 9% across different forces. A review of academic evidence suggests that between 11%-22% of adults in police custody have a relevant condition.
This significant level of unmet demand appears to have remained similar over many years. In 1993, the Royal Commission on Criminal Justice (para. 84) reported that (a) police identified the need for an AA in 4% of adult detentions; and (b) the actual level of need was 15%-20%.
Table: Results of prevalence studies of mental ill health & mental vulnerability in police custody
Failure to contact an AA
Research has shown that, in practice, police often ignore the PACE definitions and thresholds (relating to when an AA must be called) in favour of their own definitions of who requires one. Police officers do not always require an AA even where mental vulnerability is identified.
In one study, a computerised risk assessment developed by clinicians was found to make a significant difference to the rate at which mental illness and other mental vulnerability was identified by police - but made no difference to the proportion of cases for which an AA was required.
Developers should also be aware that the level of apparent demand is likely to be affected by levels of AA service provision. Data provided by forces indicates a correlation between the rate of indentification of need for an AA by police and the availabilty of AAs to the police. In areas with historically poor provision, data is likely to underestimate actual demand. Developers should consider how current levels of demand may change as a new scheme becomes available and awareness increases.
Chart: The effect of AA service provision on police identification of need (There to Help, 2015)
Local attitudes, training and policies can also have a significant effect on whether the identification of mental vulnerability results in an AA being called (as required by PACE). Early data collected by liaison and diversion pathfinder areas shows significant variation in whether people that engaged with their service also had an AA. In theory, this should be approaching 100% as L&D is primarily a mental health / vulnerabilty service.
Chart: Percentage of adults engaging with L&D who had an AA, by area [1st Sept - 31st Dec 2014] (There to Help, 2015)
Police information systems
In addition, police information systems are not designed to capture and report this data. Some electronic systems have a specific field with a ‘drop down box’ to record that an AA is required. In others, the need for an AA is captured somewhere in ‘free text’, making it very time-consuming to retrieve the data. If a manual trawl of records is required to extract data and it may be more efficient and effective to conduct a time-limited ‘snapshot’ sample in custody.
Most forces have not historically captured data on the source of the AA (i.e. whether there was demand on the organised scheme).
Most forces do not currently record information about the needs of people who are interviewed voluntarily. Specific practices are extremely varied across forces and local enquiries should be made.
Accessing data from partners
Local partnerships may have access to additional statistics on local prevalence of relevant conditions. This may include data from liaison and diversion, public health and the local joint strategic needs assessment (JSNA).
When considering prevalance data, the context of police contact should be kept in mind. For example:
- A clinical diagnosis is not required in order to meet the the PACE threshold
- Along with the fact that the prevalence of people with mental health conditions is higher in police custody than in the general population.
Developers should be aware that, in addition to prevalance data, L&D staff may hold data on the number of people who have engaged with their service who had an AA. It is important to note that this is likely to be lower than the actual level of need.
If there is an existing AA scheme in place, it should hold call-out data which should be compared to police and other partner data to identify gaps in identification.
 Dehaghani, R. (2016). He’s Just Not That Vulnerable: Exploring the Implementation of the Appropriate Adult Safeguard in Police Custody, The Howard Journal Vol 55 No 4. December 2016
Voluntary interviews can be considered to require a single AA call out. However, the situation in custody is more complicated.
There are a number of distinct phases to custody (booking in, identification, interview, charge and disposal) for which an AA should be present. There can be a significant time gap between booking-in and interview, and again between interview and charge. This will vary depending on the complexity of the case, the extent to which police choose to expedite the investigation, the availability of police resources, the need for rest periods and the length of delay while the Crown Prosecution Service make a charging decision.
A single custody episode could typically generate between one and three AA call outs. In addition to the above factors, this will also depend on when, and for how long, an AA was present. For example, if an AA attended immediately after detention and stayed until charge/bail, this would be only one call out (though it may be several hours). However, a single call out could also represent an AA arriving shortly before interview and leaving immediately afterwards.
Police information systems are unable to be able to capture and report this complexity of dat. It may not be possible to easily distinguish from the data where multiple AAs have been called for one detention episode and where multiple AAs have been called for multiple detentions. It may also be unable to differentiate between parental and scheme-based AAs.
A PACE interview (an interview held under caution) of a suspect may take place without that person being arrested and detained. Of the necessity to arrest criteria are not met, a suspect may be invited to attend an interview voluntarily.
This approach has a number of advantages, including avoiding the negative psychological impact of detention on vulnerable adults and children and reducing risk for the police.
Voluntary attendance appears to be an increasing trend in police practice. It is therefore likely that data provided by the police custody function will show a downward trend in terms of demand. When developing or commissioning a scheme, it is important to also take account of demand for voluntary interviews to avoid underestimating demand.
However, HM Inspectorate of Constabulary has noted that data collection on voluntary attendance interviews is has been limited across police forces. The National Police Chiefs Council has confirmed that current information systems do not have the capacity to record and retrieve this data.
With no historical records it will be necessary to estimate the demand. As an indication, HMIC found that across three police forces over 12 months:
- Recorded instances of voluntary attendance: 10,898
- Custody throughput: 57,170
This suggest that voluntary attendance makes up around 16% of the total instances. Therefore an estimate of voluntary interviews could be arrived at by custody throughput x 0.19. Further estimates would have to me made regarding the number of adults for whom an AA would have been required. However, the use of voluntary interviews is likely to vary significantly between forces. A snapshot survey over a fixed period may be a more reliable indicator of demand. .
The same PACE safeguards apply to voluntary interviews and appropriate adults are required for the same groups of people as in custody. However, the need for AAs in voluntary interviews may currently be being under-identified because:
- responsibility to apply the AA safeguard rests with the interviewing officer who (relative to the custody sergeant) is not independent of the investigation, likely to have less knowledge of the PACE Codes, and may have less experience in identifying vulnerability;
- it is less likely that there will be access to liaison and diversion services to support police decision-making around vulnerability in adult suspects.
 HM Inspectorate of Constabulary (2015), The welfare of vulnerable people in police custody
Organised schemes are not expected to meet all of the demand for appropriate adults.
There is no national data set on the proportion of AA callouts which are served by organised schemes.
Police information systems do not typically capture the source of an appropriate adult (e.g. parents vs organised scheme). Where this is captured, it will usually rely on custody officers inputting free text rather than dedicated database fields. This can make it difficult to retrieve the data for assessing demand.
Under PACE 1984, in the case of children parents should be the first people considered as the AA.
Youth offending teams will hold data on the number of times organised provision has been used. This can be compared to police data to establish the proportion of AAs coming from other sources.
Under PACE Code C, an experienced person is considered preferable as the AA but the person may choose a family member if they prefer.
Where an existing scheme exists, the provider and/or commissioner will hold the relevant data. This can be compared to police data to establish the proportion of AAs coming from other sources.
If there is no AA scheme currently in place, a snapshot survey can be used to collect data.
Local partnerships should seek to ensure that PACE safeguards are applied fully in voluntary interviews (as in custody). This includes:
the identification of vulnerability in adults;
the requesting of an AA;
the responsiveness of organised AA schemes to requests for voluntary interviews (including outside of police stations).