Complaints management information

Report: 10
Date: 13 June 2002
By: Commissioner

Summary

Since the Directorate of Professional Standards restructured in 2000, there has been a increased focus on performance in complaint investigations. This report demonstrates that this focus is beginning to have a positive impact. Public complaint allegations are at there lowest since 1996/97. They have reduced by 15% over the last year and at an average of an 11% reduction over the past 5 years.

In 2001/2002 there has been a substantial increase in the number of substantiated allegations as a percentage of fully investigated allegations. The percentage for 2001/2002 is 14% compared with 9% in 2000/2001. The backlog of complaints over 120 days has reduced by 10% during 2001/2002. A 'Chi-square' analysis has been conducted of the diversity information.

A. Recommendation

1. Members are asked to note the report ; and
2. Members are asked to consider the way forward in responding to the results of the statistical analysis.

B. Supporting information

1. The summary of key performance indicators is attached at Appendix 1 and focuses on timeliness, quality and outcomes.

2. The diversity information is attached at Appendix 2.

3. Since the Directorate of Professional Standards was formed in October 2000, there has been an increased focus on performance in complaint investigations. The positive effects of tighter focus are beginning to be demonstrated. This report provides a commentary to the performance against complaints for the planning year 2001/2002 (1 April /2001 - 31 March 2002).

4. In addition to the usual figures supplied with this report, additional analysis has been conducted demonstrating trends in public complaints over the last 5 years (where figures are available). The table is attached at Appendix 3.

Key performance indicators - complaints reduction

Public complaints (BVPI 21) – rows 1, 2 and 3 Appendix 1

5. The number of public complaint allegations recorded in 2001/02 has reduced by 15% compared to the previous year. The 5 year (Appendix 3) trend has seen a year on year reduction since 1997/8. The average yearly reduction is 11%.

6. The number of cases has reduced by 12% in the last year with a 5-year trend of an 11% reduction year on year.

7. The monthly figure for the percentage of substantiated allegations shows an average of 19.65% for the 6 months in Appendix 1. The figure for March is therefore in line with the 6 month average. The 2001/02 figure of 14% demonstrates an increase in the number of substantiated allegations as a percentage of fully investigated allegations (Appendix 3). The 5 year figure is averaged at 9%. The impact of the higher 2001/02 figure is to increase the percentage of substantiated allegations from total complaints from 2% to 3%. Despite the falling numbers of complaints over the 5 year period, the number of substantiated cases as a percentage, has remained constant.

Suspended officers / restricted duties – rows 4/5 Appendix 1

8. Over 2001/02 the number of officers suspended from duty has reduced significantly from 91 at the end of March 2001 to 57 at the end of March 2002. The positive impact of the reducing numbers of suspended officers has previously been reported as a Best Value saving equating to a non-cashable efficiency saving of £1.37 million a year (PSPM/02/24).

9. The numbers of officers on restricted duties has remained fairly constant over the last 6 months.

Investigative outcomes (public complaints) – row 9 Appendix 1

10. In the report to PSPM on 11 April (PSPM/02/23), comment was made regarding the number of allegations finalised in February being significantly lower than the previous month. The previous figure of 239 has now risen to 328 for February 2002. The reason for this increase is a delay in the data collection system. Where cases are submitted after the figures have been collated, re-running the data results in additional cases being entered for the month. It is now clear that originally reported figures are not symptomatic of a trend.

11. Similarly, the percentage of Immediate Informal Resolutions, as a proportion of all outcomes, has increased when compared to previously reported figures e.g. February 2002 14.2% to 17.1%. This too is due to recording delays and is not due to a downward trend in the percentage of Immediate Informal Resolutions being carried out.

12. Appendix 3 refers to the overall trends in completed allegations since 1996/1997. This indicates that the trend in the number of cases/allegations finalised each year is downward – on average by 12%. This accords with the downward trend in the number of cases/allegations being received.

13. The table in Appendix 3 demonstrates an overall reduction in the number of public complaints over the last 5 years. Examining the figures demonstrated per 100 officers shows an interesting measure of this reduction. The number of allegations per 100 officers is down from 21 in 1996/7 to 12 2001/02, whilst the number of substantiated allegations per 100 officers remains fairly constant at about 0.7.

Timeliness – rows 10/11 Appendix 1

14. During 2001/2002 the number of public complaint cases over 120 days has been reduced from 409 at the end of January 2002 to 369 in March 2002 a reduction of nearly 10%. An increased focus on timeliness issues has begun to have an impact on performance in this area.

15. Between the end of June 2001 and the end of March 2002 the total number of current public complaint investigations has also reduced significantly by 17% (1012 to 844).

16. The average number of days taken to complete a public complaint investigation falls short of the 120 target by a considerable margin. Since this target was specifically monitored, the average over the three months January to March 2002 is 282 days. However, the trend is downward.

17. The MPS is not alone in failing to meet timeliness targets. Both the Police Complaints Authority (PAC) and the Crown Prosecution Service (CPS) are in a similar position. It is possible therefore that the targets are not right. It is also recognised that investigations vary in complexity, which can impact on timeliness. The MPS is working in partnership with the PCA to agree a system identifying complexity factors in an investigation. The complexity will then be reflected in a sliding scale of timeliness targets. Work continues with the Learning Lab to identify ways of streamlining systems to improve timeliness.

18. As the total number of complaints has reduced, there has not been a similar improvement in timeliness. The reasons for this are not clear, although it is possible that investigations may be of a higher quality. The trend appears to be improving and consolidation following a period of intense change within the DPS may see further performance improvements. The situation will continue to be monitored by managers within the DPS.

Misconduct – rows 12 and 13 Appendix 1

19. The target set for making a decision about misconduct proceedings has been exceeded every month since data was first collected in December 2002.

20. The total number of papers dealt with by the unit is small. In March, only two sets of papers missed the target resulting in a large percentage failure. Both were complex cases, one requiring consultation with the PCA, which only missed the target by 5 days. The benchmark of 30 days is realistic for the majority of cases; certain complex cases will inevitably exceed the target.

21. Misconduct hearing data relating to the boardrooms is under review and does not appear for March 2002.

CPS/PCA decisions – row 14/15 Appendix 1

22. Over the period June 2001 to March 2002 there has been little change in the average time taken for the CPS and PCA to provide DPS with a decision. The figures continue to represent a considerable delay to the final resolution of the complaint. The delay has been discussed with both organisations and our performance data has been shared with them. Both have undertaken to review timeliness issues.

Diversity information

23. The diversity information template at Appendix 2 provides an overview of diversity information.

24. The revision of the employment tribunal and grievance database was a significant piece of work and is nearing completion. The database was completely re-written and the original case files were examined enabling the data to be validated from the source documents. Once it is fully operational, the figures will be able to be broken down into Police and Civil Staff.

25. The database revision will also identify the individual's ethnicity, which was not previously recorded on the system.

Diversity research

26. The detailed analysis into disproportionality is continuing. In order to make sense of the statistical picture, particularly when small absolute numbers create skewed percentage figures; the MPS Consultancy Group was commissioned to conduct a 'Chi-squared' analysis on the data in Appendix 5. The draft report is attached at Appendix 4.

Methodology

27. The Consultancy Group chose Chi-squared tests to carry out the analysis because it is appropriate to use on counts of data (e.g. the number of complaints recorded). Chi-squared calculates an 'expected' value for each count i.e. if a count was the result of chance, this is the value that would be expected. If the actual numbers are sufficiently different from this, then the result will be statistically significant.

28. If a result were found to be significant this would mean that a relationship existed between the number of recorded complaints and the ethnic groups. If a result was not significant then the numbers could be a result of chance and no interpretation of a set of data such as this should be made.

29. It should be noted that a significant result suggests that a relationship is present, but it cannot offer possible causes of that relationship. It is intended that these results will inform debate with Members and Independent Advisors from the DPS IAG.

Preliminary results - public complaints (recorded) complainants

30. The required population for this test was ‘the number and ethnicity of people who come into contact with the metropolitan police’. The population used for this test was the estimated resident population of London in 2000, as this was the only data available. This data source is not totally satisfactory, because it is based on the assumption that contact with the police is proportional to the resident population, which is unlikely to be the case. For this reason, caution needs to be applied when considering the results of the test.

31. The result of this test was both valid and significant. This means that the data used showed that a relationship existed between the number of public complaints and the ethnic group. A far higher number of complaints recorded were from black members of the public than would be expected if the numbers were a result of chance. A graph is attached as an Appendix 5 to visually represent the results.

32. This analysis provides a good start point to further in-depth academic work. The chi-square test provides the questions to ask rather than the answers.

Public Complaints (Recorded) Officers

33. The population used for this was the number of officers employed by the MPS during 2001/02 according to ethnic group. It should be noted that police officer populations are a measure of all officers and not just those that come into contact with the public. There may be different proportions of officers in non-operational posts, and this has the potential to offer a result that is not consistent with the true picture.

34. The result of this test was significant. This means that the data used showed that a relationship existed between the number of public complaints and the ethnic group.

35. A higher number of complaints were recorded of both black and Asian officers than would be expected if the numbers were a result of chance.

'Stop and Search only (recorded)’

36. Only ten months data relating to the numbers and ethnicity of people stopped & searched for 2001/02 was available for this analysis. The final data was extrapolated using mean averages. Recognised as being crude, this test was run using both 2000/01 and 2001/02 data.

37. The population used was the number of people stopped and searched during the relevant time period according to ethnicity.

38. The results for both these years were significant. This means that the data used showed that a relationship existed between the number of complaints made, as a result of stops and searches, and the ethnic group.

39. A lower number of complaints were received from white complainants than expected, and a higher number of complaints were received from both black and other ethnic complainants than would be expected if the numbers were a result of chance.

‘Racial discriminatory behaviour allegations only (recorded)’

40. The same population was used as in public complaints and therefore has the same cautionary note attached.

41. The test was significant, a far higher number of allegations of discriminatory behaviour recorded were from black members of the public than would be expected if the numbers were a result of chance. This is probably not an unexpected result for this category of complaint.

Internal investigations (recorded) officers

42. The result of this test was not significant, meaning there was no association found between the number of internal investigations and the ethnic groups. This seems to indicate that disproportionality is not a factor in internal investigations. The situation will, however, continue to be monitored.

Suspended officers

43. Although this test was run and found to be significant, the absolute numbers involved were very small making the drawing of conclusions regarding disproportionality difficult.

Written Warnings

44. Again the numbers were small, although the test was run and found to be not significant.

'Informal discipline as a result of investigation (Public complaint - excluding written warnings)’

45. In order to validate the test, three years’ data was used with 'Other ethnicity' removed. The result from this test was valid and also found to be significant. This means that the data used showed that a relationship existed between the number of informal discipline (public complaints) and the ethnic group.

46. A higher number of informal discipline proceedings were recorded of black officers and lower number of were recorded of Asian officers than would be expected if the numbers were a result of chance.

‘Informal discipline as a result of investigation'  (Internal investigations - excluding written warnings)

47. In order to validate the test, three years’ data was used with 'Other ethnicity' removed. The result from this test was valid and also found to be significant. This means that the data used showed that a relationship existed between the number of informal discipline (internal investigations) and the ethnic group.

48. A higher number of informal discipline proceedings were recorded of both black officers and Asian officers than would be expected if the numbers were a result of chance. A slightly higher number would be expected of white officers.

‘Misconduct/discipline hearings (public complaints and internal investigations)'

49. In order to validate the test, it was necessary to collapse the table to contain just ‘white’ and ‘visibly ethnic’ categories.

50. The result of this test was valid but not significant, meaning there was no association found between the number of hearings and the ethnic groups. Any differences could be the result of chance.

'Dismissals or required resignation as a result of misconduct/discipline hearings (public complaints)'

51. The numbers involved were so small, that the test was not run.

'Dismissals or required to resign sanctions as a result of misconduct/discipline hearings (internal investigations)

52. In order to validate the test, it was necessary to collapse the table to contain just ‘white’ and ‘visibly ethnic’ categories. The test was not significant.

Commissioner's appeal/AC review (public complaints)

53. The numbers involved were so small, that the test was not run.

Commissioner's appeal/AC review (internal investigations)'

54. In order to validate the test, it was necessary to collapse the table to contain just ‘white’ and ‘visibly ethnic’ categories. The test was not significant.

Secretary of State/police appeals tribunal (public complaints and Internal investigations)

55. The numbers involved were so small, that the test was not run.

Grievances

56. In order to validate the test, three years’ data was used. The result from this test was valid and also found to be significant. This means that the data used showed that a relationship existed between the number of grievances and the ethnic group.

57. A higher number of grievances were recorded from black, Asian and Other ethnicity officers than would be expected if the numbers were a result of chance. A lower number was recorded than would be expected of white officers.

58. The grievance procedure is being revised and re-launched. It is likely that this will encourage greater use of internal means to resolve disputes will see an increase in the numbers of grievance procedures being lodged. This is likely to be reinforced by proposed legislative changes which will emphasise the expectation that employees will explore internal resolution avenues before going to litigation unless there are compelling reasons why this is inappropriate. These measures, if effective, should lead to a reduction in the number of cases going to employment tribunal.

Employment Tribunals

59. In order to validate the test, it was necessary to collapse the table to contain just ‘white’ and ‘visibly ethnic’ categories.

60. The result from this test was found to be significant. This means that the data used showed that a relationship existed between the number of employment tribunals and the ethnic group. A considerably lower number of white officers went to employment tribunals than would be expected by chance, the reverse was found to be true of visibly ethnic officers. A graph is attached as an appendix to the report visually demonstrating the results.

61. Disproportionality should be expected in Employment Tribunals as in the absence of contracts of employment for police officers, the only grounds for bringing an employment tribunal, are for sexual or racial discrimination.

62. Section 4 of the report at Appendix 4 responds to the request from DPS to give a view as to whether another statistical tool for calculating standard deviations from the mean could be used to analyse the data. Unfortunately, the test is not appropriate for this type of data.

The way forward

63. The analysis conducted so far together with further research where appropriate will be discussed with the MPS Independent Advisory Group and staff association representatives. MPA Members may also wish to consider how they can further contribute to this debate.

64. Appropriate action plans will be drawn up which could cover issues such as awareness of diversity issues in complaints, promotion of the use of the alternatives to the formal discipline such as wider use of the Unsatisfactory Performance Procedure.

C. Financial implications

None.

D. Background papers

None.

E. Contact details

Report authors: Detective Superintendent Chris Bourlet, and HEO Michael Clark, MPS.

For information contact:

MPA general: 020 7202 0202
Media enquiries: 020 7202 0217/18

Send an e-mail linking to this page

Feedback