1
00:00:00,420 --> 00:00:09,450
And both these projects were collaborative projects. And I was collaborating with Chris, Jack and Antonio from around Europe on one of these projects.

2
00:00:09,690 --> 00:00:17,130
Chris used to be on the Dphil students here. He left about a year or so ago and we started this project while he was still on his dphil here.

3
00:00:17,340 --> 00:00:21,570
And I have a collaborator on the on the other experiment. That's Sara McQuaid from the University of Edinburgh.

4
00:00:21,720 --> 00:00:27,690
And I have to say, in both cases I did most of the work. So in a sense, I'm taking no glory is not the right.

5
00:00:27,690 --> 00:00:30,990
Well, I'm taking credit for actually a huge amount of work done by other people.

6
00:00:31,390 --> 00:00:36,090
Some of this, anyway, experiments in police confidence and police legitimacy.

7
00:00:36,270 --> 00:00:40,139
I'm going to start off with a bit of what could catch us is deep background to this.

8
00:00:40,140 --> 00:00:46,170
Just just framing this in a couple of debates within the subject facing world peace studies, much white,

9
00:00:46,170 --> 00:00:52,710
much more widely and then a bit more immediate background as to why do we care about these issues of public confidence and place such as this,

10
00:00:52,980 --> 00:00:56,459
particularly in the context of Great Britain experiments?

11
00:00:56,460 --> 00:01:00,150
And we're talking about what took place in England, the other took place in Scotland,

12
00:01:00,240 --> 00:01:05,640
and it's at least argue that something specific about these contexts which make these kinds of exponents particularly relevant.

13
00:01:05,850 --> 00:01:10,589
And is it also arguable that they may not travel particularly well to have a policing context?

14
00:01:10,590 --> 00:01:12,330
I think that's something worthy of discussion.

15
00:01:12,630 --> 00:01:18,240
Later on, I'm going to suggest that in Pursuit of Justice, which directly motivated one of the experiments,

16
00:01:18,420 --> 00:01:20,700
I think a bit about the relationship between police and community,

17
00:01:20,700 --> 00:01:27,509
the pursuit of justice model and overrides kind of themselves, motivates and then would describe the results of two experiments.

18
00:01:27,510 --> 00:01:37,170
And one one was a quasi experiment on the effects of mounted police community patrols in three areas in Gloucestershire, England and in London.

19
00:01:37,440 --> 00:01:43,500
And the other is a replication of a quite well known study from Australia to Queensland community engagement trial.

20
00:01:43,710 --> 00:01:47,250
Q Sets that we conducted in Scotland, that's obviously the project I review.

21
00:01:47,250 --> 00:01:51,659
Sarah McQueen I'm going to finish with some questions you have noticed I've lost from the title.

22
00:01:51,660 --> 00:01:58,410
So the mix about a comment I had in the title of which many promises and pitfalls as many because I'm not really

23
00:01:58,410 --> 00:02:03,570
qualified and I'm not that strong methodology to talk about the experimental method quite in quite that way.

24
00:02:04,030 --> 00:02:10,019
I think some of the promises and pitfalls of the use of experiments in criminological contexts come out in the course of discussion.

25
00:02:10,020 --> 00:02:15,329
Or at least I hope they do. So the first bit of background is this, and this really is deep background, I think.

26
00:02:15,330 --> 00:02:21,840
I think what I'm talking about today, the kind of research I'm talking about today, really presupposes two things.

27
00:02:21,990 --> 00:02:28,649
It presupposes that people want and at least expect at some point in the future to get democratic policing.

28
00:02:28,650 --> 00:02:32,040
Obviously, that defined democratic policing in lots of lots of ways.

29
00:02:32,310 --> 00:02:38,910
Most of the definitions condense around something like this. So democratic police organisations have respect for citizens rights.

30
00:02:39,090 --> 00:02:44,820
They use procedural fairness, they used state of justice, they use the minimum or appropriate levels of force.

31
00:02:45,390 --> 00:02:50,580
They're accountable. They're responsible to those they serve. They engage in citizen participation.

32
00:02:51,030 --> 00:02:56,280
They operate in equitable ways. They're responsive to the needs of those they serve, etc., etc.

33
00:02:56,460 --> 00:03:01,170
I really very least would presupposing that people want and I think it's feasible they can get this,

34
00:03:01,410 --> 00:03:06,840
I think it's also possible we're presupposing that policing actually is like this, at least in some context.

35
00:03:06,840 --> 00:03:12,060
For example, within the United Kingdom, the British police do meet some of these criteria, at least some of the time.

36
00:03:12,240 --> 00:03:17,430
And if we were dealing with a police force in a different situation where the police find it very difficult to meet these criteria,

37
00:03:17,640 --> 00:03:21,510
I think a lot of what I'm saying is they might be problematic, to say the least.

38
00:03:21,750 --> 00:03:25,590
That's the first bit background. The second bit of background really leads on from that in a sense,

39
00:03:25,590 --> 00:03:30,210
and that's the image of the imaginary of the ideology that surrounds policing in Great Britain.

40
00:03:30,450 --> 00:03:34,920
And obviously that was once interesting in Great Britain and the United Kingdom for obvious reasons.

41
00:03:35,220 --> 00:03:40,020
And this is one set of images you could construct around policing in this context.

42
00:03:40,020 --> 00:03:44,040
I think then the key, if you can see that that key message is meet your neighbours.

43
00:03:44,250 --> 00:03:46,620
So we were operating in a context where people can imagine,

44
00:03:46,620 --> 00:03:51,749
at least imagine that the police can sometimes be their neighbours and it makes sense to kind of the police could be a

45
00:03:51,750 --> 00:03:56,940
neighbours that they have something they want to have things to do that want to engage if they want to act on your priorities,

46
00:03:57,120 --> 00:04:03,779
etc., etc. I could go on, but I'm sure you get the message. And of course that's not the only image of policing that exists in the United Kingdom.

47
00:04:03,780 --> 00:04:11,130
Here's another set of images. These things run alongside each other always and interact with each other in complicated ways.

48
00:04:11,730 --> 00:04:17,880
But nevertheless, I think it's true that there's a particular cultural resonance of policing that persists in England, Wales and Scotland.

49
00:04:18,060 --> 00:04:22,200
And a lot of what I'm saying today is a sense founded on the idea that people still,

50
00:04:22,200 --> 00:04:30,270
still do think about policing in something like these tanks, at least in sometimes in some places within this particular social context.

51
00:04:31,500 --> 00:04:38,530
The real background out of the background motivating these studies is really a trend in both criminology

52
00:04:38,530 --> 00:04:43,380
and in both policing studies and in policy circles and within police organisations themselves.

53
00:04:43,530 --> 00:04:48,540
That you could say started in the 1990 that I think you could also said it started a long time before that,

54
00:04:48,810 --> 00:04:54,270
and that was a specific set of concerns arose about declining public trust and confidence in the police.

55
00:04:54,540 --> 00:04:59,910
And this really started to come onto the policy agenda in the late 1990s into the 2000, obviously reacted.

56
00:05:00,120 --> 00:05:02,850
Two events around the time, for example, the Steve Lawrence inquiry,

57
00:05:03,030 --> 00:05:07,500
but also reacting to long term historical trends are already starting to be identified.

58
00:05:08,070 --> 00:05:11,520
At that time, this led to many, many different policy interventions.

59
00:05:11,730 --> 00:05:14,400
We can think about reassurance, policing, neighbourhood policing, etc.,

60
00:05:14,400 --> 00:05:22,469
etc. and it ultimately led to the establishment by the last Labour government of a single overarching target performance target for police in England.

61
00:05:22,470 --> 00:05:24,870
Wales, which was a measure of public trust and confidence.

62
00:05:25,080 --> 00:05:31,470
And police organisations were jointly and severally tasked to increase increasing public confidence across a variety of measures.

63
00:05:32,280 --> 00:05:37,169
This obviously and perhaps not surprising access expert research from the WHO is now saying, if you like,

64
00:05:37,170 --> 00:05:42,819
if you like another stage of this research that started in the 1990s and this work at its end

65
00:05:42,820 --> 00:05:47,580
that this this concern at its outset was based around in essentially one survey question.

66
00:05:47,820 --> 00:05:52,320
The survey question has been in British conservation since inception, and it's the famous question asked,

67
00:05:52,320 --> 00:05:57,690
people taking everything into account how good a job do you think the police in your local area are doing?

68
00:05:57,990 --> 00:06:03,210
And if we look at the trends from the British Crime Survey, this is from 1984 to 2005 six.

69
00:06:03,360 --> 00:06:09,480
I think you could conclude on that basis that reports of a haemorrhage of public support for police are overstated.

70
00:06:09,490 --> 00:06:12,930
You will see this language in some of the literature around this. This didn't happen.

71
00:06:13,140 --> 00:06:16,980
So these bars here, if you can see this, the bar is the people rating police.

72
00:06:16,980 --> 00:06:21,330
Very good. The middle part is fairly good, fairly poor.

73
00:06:21,570 --> 00:06:28,340
Very poor. So most of the concern was around the decline in the proportion of people rating the police is very good stuff.

74
00:06:28,340 --> 00:06:32,790
Talking about 30% in 1984 to about 50% in 2003 four.

75
00:06:32,940 --> 00:06:37,080
It changed the question, what is it, 2000? Pretty slightly. No, it doesn't matter too much about that.

76
00:06:37,170 --> 00:06:42,390
And at the same time, there was a decline in this very, very strong, very, very good rating.

77
00:06:43,080 --> 00:06:46,650
The fairly poor and very poor ratings increase, but it's not huge.

78
00:06:46,770 --> 00:06:51,660
You might look at this chart and look down at the motivations for the kind of work I'm talking about today.

79
00:06:51,780 --> 00:06:56,460
I think you'd be mistaken to do that. And I put two charts in mainly because I just really like them.

80
00:06:56,550 --> 00:07:03,090
I think they illustrate the point I'm trying to make. So in this chart here on the horizontal axis we've got yes, on the vertical axis,

81
00:07:03,090 --> 00:07:06,660
we've got proportion of people rating their local police as very good.

82
00:07:06,870 --> 00:07:10,620
And then each one of those lines represents the views of people born in a particular decade.

83
00:07:10,860 --> 00:07:15,900
So the blue line at the top, the dark blue line is people born in the night before the 1920s.

84
00:07:16,170 --> 00:07:22,740
Next slide down as people born in the 1920s, in the 1930s and Nazi forces, etc., etc. just to kind of suit a cohort analysis.

85
00:07:22,950 --> 00:07:24,870
And what it clearly shows in the early years,

86
00:07:24,880 --> 00:07:30,750
a huge proportions of people born quite a long time ago who were writing their local police as very good.

87
00:07:31,050 --> 00:07:39,450
But as these two decades went on, their views collapsed down to meet the views of the younger people who'd always been slightly more suspicious,

88
00:07:39,450 --> 00:07:42,720
slightly more sceptical of the police. This is real funnel shape.

89
00:07:42,870 --> 00:07:48,090
So by the time we get to the most recent year, there's almost no variation in age and opinions of the police.

90
00:07:49,350 --> 00:07:52,740
The views of the people who previously had the most favourable perception of the police

91
00:07:52,860 --> 00:07:57,179
had gone down to meet those who previously had the least favourable this chart.

92
00:07:57,180 --> 00:07:59,610
So the same story, just exactly the same thing in a sense.

93
00:07:59,760 --> 00:08:05,579
But now we have the views of a notional 50 year old man, white man, living outside the inner city.

94
00:08:05,580 --> 00:08:11,730
That's the blue line. And the red line is the views of a notional 20 year old man from an ethnic minority group living in the inner city.

95
00:08:11,850 --> 00:08:14,120
And again, you've got this funny shape so often.

96
00:08:14,190 --> 00:08:19,500
What I think both these charts demonstrate is there really was a decline in trust and confidence over this period.

97
00:08:19,620 --> 00:08:24,900
It really was something that people needed to be concerned about and they were right to pick up on as an issue.

98
00:08:25,110 --> 00:08:31,860
It wasn't of overweening importance, which is some of the things in some ways it's presented in various parts of the literature,

99
00:08:32,850 --> 00:08:35,690
and you're not necessarily observant among you.

100
00:08:35,690 --> 00:08:42,870
Even if you noticed, I put scare marks around trust and confidence throughout so far as a number of reasons for doing this.

101
00:08:42,870 --> 00:08:48,900
I mean, this is the commonly accepted term, which seems to be the commonly accepted term for talking about public opinion and police.

102
00:08:49,050 --> 00:08:54,440
People will always talk about trust and confidence. No one ever sat down to think what they meant by this mean.

103
00:08:54,550 --> 00:08:57,959
I think I'll go on to suggest to the minute that it's quite a lot to do with legitimacy.

104
00:08:57,960 --> 00:09:02,250
But people didn't really want to talk about declining legitimacy. The police, that sounded a bit scary.

105
00:09:02,430 --> 00:09:06,209
So they said trust and confidence and said no one ever worked out what to trust.

106
00:09:06,210 --> 00:09:14,550
The police might actually me I'm not to any great extent and not so that facing up to this

107
00:09:14,940 --> 00:09:18,270
various police organisations start to say we need to know more about public opinion,

108
00:09:18,390 --> 00:09:25,230
we need to unpack this trust and confidence to find out what's going on underneath it and what factors might, in a sense, drive trust and confidence.

109
00:09:25,440 --> 00:09:32,670
So to fascinate 2009, I had some colleagues from I to see started to do some work with the M.P.s to generate a kind of more robust,

110
00:09:32,880 --> 00:09:36,360
more robust set of measures around the issues of trust and confidence,

111
00:09:37,050 --> 00:09:42,180
which really got to grips with what might be going on underneath and really engage with the literature on trust, for example.

112
00:09:42,660 --> 00:09:45,810
And the NPF still has targets based on these measures to stay safe.

113
00:09:45,820 --> 00:09:53,400
Our commanders in the Met are still tasked with increasing public trust and confidence across some of the measures that I'm just about to describe.

114
00:09:53,880 --> 00:09:59,820
And I've seen the literature on trust is vast and we can only ever look at small amounts of it, I think.

115
00:09:59,930 --> 00:10:03,410
Titular relevant definitions are understandings of trust.

116
00:10:04,010 --> 00:10:07,700
In this context, revolve around the work of people like Baba and Hardin.

117
00:10:07,910 --> 00:10:14,690
So Baba says that trust involves expectations that those we trust will be technically competent to do the things we are trusting them to do,

118
00:10:14,960 --> 00:10:20,510
and that those we trust will place our own interests above our own. And of course, that therefore implies that they know what our own interests are.

119
00:10:20,630 --> 00:10:24,800
They understand what our interests are within a particular trust relationship.

120
00:10:25,430 --> 00:10:29,719
Part is, says the same thing slightly more succinctly to say we trust you means we believe you have

121
00:10:29,720 --> 00:10:33,830
the right intentions towards us and you are competent to do what we trust you to do.

122
00:10:34,040 --> 00:10:40,970
Just about intentions. It's about competency. I think you could also say it's about expectations and evaluations.

123
00:10:41,150 --> 00:10:44,060
When you trust someone, you expect them to behave in a certain way.

124
00:10:44,270 --> 00:10:49,700
But your expectations, at least partly based on your evaluation of their prior behaviour or performance.

125
00:10:49,850 --> 00:10:53,330
So trust always wraps up, I think, expectation and evaluation.

126
00:10:53,600 --> 00:10:59,900
And the other thing about trust in this context and indeed mothering actually talks about trust in a business context more than anything else.

127
00:11:00,080 --> 00:11:06,050
He says trust always involves a leap of faith. You're always trusting in relatively low information context.

128
00:11:06,170 --> 00:11:12,080
If you had all the information about someone, you wouldn't need to trust them because you'd know what they were going to do in a particular context.

129
00:11:12,170 --> 00:11:18,950
And trust also is also therefore always implies it implies contexts of risk.

130
00:11:19,760 --> 00:11:24,530
So based on this work and based on some other work that people don't past, obviously we didn't come to this on our own, really.

131
00:11:25,220 --> 00:11:28,020
We define three dimensions of trust in the lens Metropolitan Police,

132
00:11:28,020 --> 00:11:31,650
but I think these could apply to police organisations very much more widely in that.

133
00:11:31,790 --> 00:11:35,780
So it's often a slice of fun, a bit wrong with us to talk about trust in police effectiveness,

134
00:11:36,110 --> 00:11:39,440
trust in police community engagement and trust in police fairness.

135
00:11:39,680 --> 00:11:43,549
Trust the police effectiveness as a competence. Trust in police community engagement.

136
00:11:43,550 --> 00:11:50,000
Trust in police fairness wraps up intentions, right intentions, shared interests, motivations, so forth.

137
00:11:50,120 --> 00:11:55,260
So and we tend to find and I'm going to talk about some of the measures we use to look at these things in a minute.

138
00:11:55,370 --> 00:11:59,810
We tend to find these are empirically distinct constructs or components of trust,

139
00:11:59,960 --> 00:12:05,360
but they also sense some sense of some together to generate an overall measure of trust and confidence.

140
00:12:05,540 --> 00:12:09,670
Should you wish to do that? Why is distrust important?

141
00:12:09,680 --> 00:12:12,110
While I don't think I really need to stress this in this context,

142
00:12:12,110 --> 00:12:17,720
but there's a lot of evidence to suggest to the extent that people trust organisations or institutions such as the police,

143
00:12:17,960 --> 00:12:19,310
they're more likely to cooperate with them,

144
00:12:19,310 --> 00:12:24,950
they're more likely to defer to them, they're more likely to comply with orders and more likely to come forward of information,

145
00:12:25,100 --> 00:12:31,250
etc. There's almost the obvious link between trust and confidence and legitimacy with files already addressed.

146
00:12:32,330 --> 00:12:36,170
But there's some questions here, of course, and this is one of the reasons why they're so interesting,

147
00:12:36,170 --> 00:12:38,720
is were they're interested in not just knowing what trust looked like,

148
00:12:38,870 --> 00:12:44,360
but the kind of factors that shaped trust in the police because they've been tasked with increasing public trust and confidence,

149
00:12:44,480 --> 00:12:51,080
they wanted to know, what can we do about this? How can we how can we work with the public to enhance their sense of trust in us?

150
00:12:51,200 --> 00:12:56,809
As an organisation, we need to think a bit more about the relationship between trust and legitimacy and

151
00:12:56,810 --> 00:13:00,470
we need to make a bit more about what the police can do to bolster these things.

152
00:13:01,400 --> 00:13:08,180
This is where procedure justice theory comes in. I think this is the basis for almost all of the work I've done so far in this office.

153
00:13:08,180 --> 00:13:13,190
I think the most robust and most empirically well-supported way of understanding the nature of

154
00:13:13,190 --> 00:13:18,170
police community relations and the types of things police officers can do to enhance or of course,

155
00:13:18,380 --> 00:13:22,280
undermine trust, I should always say. And Mary's my hair because she was picking up on this.

156
00:13:23,060 --> 00:13:28,070
I always talk about these things in a very positive sense. I'm always talking about things that police can do to increase trust.

157
00:13:28,310 --> 00:13:31,280
But they all the relationships and we're talking about still have bi directional.

158
00:13:31,580 --> 00:13:36,620
They feel lots of things that police can do to motivate distrust or damage trust.

159
00:13:36,710 --> 00:13:39,680
And these tend to be the reverse. The kind of things I'm thinking about today.

160
00:13:39,680 --> 00:13:47,180
It's important to recognise that pursuit of justice theory developed by Tom Taylor and colleagues working in the United States in the 1980s,

161
00:13:47,180 --> 00:13:51,170
finding increasing purchase in criminal justice contexts around the world.

162
00:13:51,410 --> 00:13:53,450
Although it's probably not a universal. We don't think that.

163
00:13:53,450 --> 00:13:59,930
I don't think and this is really at its base a theory or a model for understanding social relations in group settings,

164
00:14:00,200 --> 00:14:04,160
particularly groups where one member of the group represents the group as a whole,

165
00:14:04,340 --> 00:14:09,140
and particularly group settings where one member of the group has significant power over other members of the group.

166
00:14:09,290 --> 00:14:15,830
I can't think I think of a better way of describing the relationship that people in person have with the police on that basis.

167
00:14:16,610 --> 00:14:21,380
It provides a way of understanding, as I say, what types of police behaviour motivate trust,

168
00:14:21,470 --> 00:14:25,340
generate legitimacy, generate cooperation, generate compliance.

169
00:14:25,670 --> 00:14:32,809
Most of you will have heard this before. So with a history of this, I'd say Justice Very says that when police, police officers treat people unjustly,

170
00:14:32,810 --> 00:14:36,889
respectfully, with dignified treatment, when they use less transparent,

171
00:14:36,890 --> 00:14:39,320
fair decision making, when they allow people to voice,

172
00:14:39,470 --> 00:14:43,190
when they look like they're listening to people's voice and nobody is prepared to act on what people.

173
00:14:43,190 --> 00:14:47,360
It's an incident that this generates a sense that the police are behaving in a fair way.

174
00:14:47,540 --> 00:14:51,199
It generates precisely a sense of trust and a sense of pursuit of justice.

175
00:14:51,200 --> 00:14:53,420
A sense of trust flows into legitimacy.

176
00:14:53,570 --> 00:14:59,600
And, of course, the pursuit of justice, moral legitimacy flows into cooperation, deference, compliance with the law, all this good stuff.

177
00:14:59,820 --> 00:15:05,430
If you like this matter flow out the hand. And of course, that is the one of the most important aspects of SEO.

178
00:15:05,430 --> 00:15:11,850
Just as there is, it says that procedural justice is a more important way of generating trust and generating legitimacy.

179
00:15:11,940 --> 00:15:15,660
Then the instruments of effectiveness of the police. When people encounter police officers,

180
00:15:15,840 --> 00:15:20,100
they tend on average to care much more about the way this police officers behave in

181
00:15:20,100 --> 00:15:24,240
relation to them than the kind of the rewards or dis benefits as police officers,

182
00:15:24,250 --> 00:15:30,960
an officer in terms of getting a seven goes back, in terms of arresting them, what those instrumental outcomes might be.

183
00:15:31,860 --> 00:15:35,189
Why is prestige justice work? Well, it works because we're dealing with group settings.

184
00:15:35,190 --> 00:15:42,870
So the basic idea is that police officers represent an important sense, a social groups which most people feel they belong, nation state community.

185
00:15:42,870 --> 00:15:44,370
So you can catch those in different ways.

186
00:15:44,580 --> 00:15:50,970
When police officers interact with people that send an important message, messages about their status, their inclusion, their value within this group.

187
00:15:51,210 --> 00:15:55,920
So fair treatment communicates you're included in this group. Unfair treatment communicates you.

188
00:15:56,040 --> 00:16:02,219
You're excluded from this group. When you feel included, you're motivated to legitimise group authorities, you're motivated to cooperate with.

189
00:16:02,220 --> 00:16:07,650
And on behalf of the group, you activists comply with its laws. So there's an important effective element procedure.

190
00:16:07,800 --> 00:16:11,490
It makes you feel excluded and that motivates a whole set of behaviours.

191
00:16:12,150 --> 00:16:17,910
I think there's also that kind of a portion evaluative aspects of procedural justice and police are exemplars

192
00:16:17,910 --> 00:16:23,040
of good comes at or at least they should do so when police officers behave in a certain way to use towards us,

193
00:16:23,160 --> 00:16:27,360
we use this to provide information. So that is this police officer behaving the way that they should behaving.

194
00:16:27,570 --> 00:16:35,130
And when they are, that motivates a reciprocal set of reciprocal sets of behaviours on our part when they're behaving in the wrong way.

195
00:16:35,280 --> 00:16:39,560
That might most fair, quite different set of behaviours on our part and of course in procedures,

196
00:16:39,570 --> 00:16:45,149
justice works and there's a lot of evidence that it does and cooperation and compliance with the police,

197
00:16:45,150 --> 00:16:49,020
with the law to be secured most importantly by process, by staff,

198
00:16:49,020 --> 00:16:54,750
is a policing that tends the relationship between police and officials, between police and community.

199
00:16:54,900 --> 00:16:58,380
And this is where the two experiments really come in, I think.

200
00:16:59,370 --> 00:17:02,519
And a quick word on defining which is missing out in terms of I think it's always

201
00:17:02,520 --> 00:17:06,660
useful to define terms and there's lots of fuzzy social science terms in all of this.

202
00:17:07,130 --> 00:17:13,800
When I'm talking about legitimacy for the purposes of today, I'm drawing most important in the work of David Bassam, who's political scientist.

203
00:17:14,040 --> 00:17:20,700
And he says that legitimacy is always granted by individuals to those who have power over them in some context,

204
00:17:20,880 --> 00:17:26,910
on the basis of common shared values, most importantly, but more specifically, three dimensions of a relationship,

205
00:17:26,910 --> 00:17:31,559
if you like, must be fulfilled before power can be considered legitimate within a particular

206
00:17:31,560 --> 00:17:36,000
social context that's going to conformity of that power authority to a set of rules,

207
00:17:36,870 --> 00:17:40,140
the justify ability of those rules in terms of shared place.

208
00:17:40,230 --> 00:17:46,730
And it was not only must the police abide by a set of rules, but we, as the police must declare in some sense the right sets of rules.

209
00:17:47,070 --> 00:17:52,200
And finally, it's not enough claims be for for people to believe the authorities are legitimate.

210
00:17:52,290 --> 00:17:58,170
They have to act in ways that both serve to recognise this legitimacy and I think empirically reproduce it on the ground.

211
00:17:58,290 --> 00:18:01,350
So he talks about expressed consent of first governed by a power.

212
00:18:01,530 --> 00:18:04,800
In the context of policing we tend to think about the duty to obey.

213
00:18:04,980 --> 00:18:09,210
We think to is interesting that the legitimate said police persist consists at least in part.

214
00:18:09,390 --> 00:18:13,110
I mean people sense that they have a duty to obey the instructions of police officer.

215
00:18:13,230 --> 00:18:16,290
And of course out there in the real world, actual acts of deference,

216
00:18:16,290 --> 00:18:22,170
actual acts of obedience to the will of these officers serves to reproduce the legitimacy of police in particular contexts.

217
00:18:23,190 --> 00:18:26,879
Primarily, what I'm talking about here, of course, is legitimacy is a psychological state.

218
00:18:26,880 --> 00:18:32,580
So legitimacy of police as it exists inside people's minds as the primary objects of concern today.

219
00:18:32,730 --> 00:18:38,430
So you can think about this again as being the recognition and justification of the right of a power to exercise that power,

220
00:18:38,550 --> 00:18:41,040
exercise influence on a group of people.

221
00:18:42,370 --> 00:18:47,250
And when you're thinking about sources of trust and legitimacy and now, as I said, I'm going to get to the experiments in a minute.

222
00:18:47,490 --> 00:18:49,979
And of course, it's not all about procedural justice.

223
00:18:49,980 --> 00:18:56,070
That would be not naive and ridiculous to say that it's not even all about effectiveness and the other aspects of police activity.

224
00:18:56,220 --> 00:19:02,250
So you might want to think, for example, about wider experiences of crime and the effect that they have may have on trust and legitimacy.

225
00:19:02,370 --> 00:19:07,739
We may want to think about symbolic aspects of policing. We may want to think about the extent to which police represent any body.

226
00:19:07,740 --> 00:19:13,660
Order and legitimacy may be granted or withdrawn on the basis that they managed to do so successfully.

227
00:19:13,660 --> 00:19:21,570
They appear to manage to do so successfully. There's a consistent link, for example, between police visibility, trust and legitimacy.

228
00:19:21,750 --> 00:19:24,150
We find that in many, many cross sectional survey contexts,

229
00:19:24,300 --> 00:19:30,480
many seeing police is in the British context associated with high levels of trust, high levels of legitimacy.

230
00:19:30,630 --> 00:19:36,540
And of course, you could also then put in a whole range of psychological factors, such as psychological factors, such as witchcraft, etc., etc.

231
00:19:36,870 --> 00:19:40,499
So you could come up with a model of trust and legitimacy, which is useful to me.

232
00:19:40,500 --> 00:19:44,850
But let's say to those of you at the back that basically all this is saying is on the left hand side,

233
00:19:45,390 --> 00:19:51,420
we've got lots of factors, lots of things that could feed into trust and just say visibility, personal contact.

234
00:19:51,420 --> 00:19:53,340
That's what I'm concentrating on today.

235
00:19:53,520 --> 00:19:59,040
Vicarious contacts of the kind of stories that circulate within families, within social groups about the police.

236
00:19:59,380 --> 00:20:05,110
Media, people's cultural repertoire and experiences of crime victimisation, expensive antisocial behaviour.

237
00:20:05,230 --> 00:20:07,120
And then we cross social psychological factors.

238
00:20:07,390 --> 00:20:13,360
So for example, authoritarianism, for example, system justification theory, for example, parties and such.

239
00:20:13,600 --> 00:20:19,660
All these things going to feed into trust and in various complicated ways and even can to mistake its legitimacy.

240
00:20:19,810 --> 00:20:25,030
And then on. In cooperation and compliance. Some questions.

241
00:20:25,030 --> 00:20:28,459
I've already addressed at least one of these. I think it's worth raising.

242
00:20:28,460 --> 00:20:33,420
You can see of motivates experiments. The first question of the first issue is most of the evidence on these issues.

243
00:20:33,430 --> 00:20:38,649
Almost everything I've said so far today is evidenced and is evidenced by survey work,

244
00:20:38,650 --> 00:20:42,610
but it's evidence almost entirely by cross-sectional survey work snapshot surveys

245
00:20:42,820 --> 00:20:46,480
that any measure of opinions and attitudes and orientations at one point in time.

246
00:20:46,540 --> 00:20:49,240
In other words, the evidence is good, but it's not that good.

247
00:20:49,390 --> 00:20:54,670
And this is where the field experiments come in, because it allows us to test whether the use of procedural fairness,

248
00:20:54,880 --> 00:20:57,910
the way that police officers handle contacts between members of the public,

249
00:20:58,180 --> 00:21:03,520
over the service members of the public, whether visibility, a causal dealings with trust and legitimacy.

250
00:21:05,410 --> 00:21:09,040
And we've also had the police per person experience contacts,

251
00:21:09,580 --> 00:21:16,810
other forms of police activity and behaviours during that activity are likely to be important generators of trust and legitimacy.

252
00:21:16,930 --> 00:21:18,390
But what can police do about this?

253
00:21:18,680 --> 00:21:24,790
What are the policies that police officers can put in place or police managers can put in place to enhance trust, enhance legitimacy?

254
00:21:24,970 --> 00:21:30,070
And one of the good things I think about doing experiments in this context is it really makes you think about this question.

255
00:21:30,340 --> 00:21:33,460
It really makes you think, okay, well, what can police do in this context?

256
00:21:33,520 --> 00:21:38,410
And thinking about what police can what can I please do in this context becomes your intervention in the experiment.

257
00:21:38,980 --> 00:21:44,620
So that's a very, very long winded setup and a bit of background not to go on to the experiments themselves.

258
00:21:44,800 --> 00:21:47,290
So I'm going to start with the matter of Community Patrol experiment.

259
00:21:47,440 --> 00:21:54,340
This was in essence, I designed a set up to look at the effects of police visibility on public trust and confidence.

260
00:21:54,490 --> 00:21:57,220
Actually, it sounds a bit more than that, and we'll come to that in a minute.

261
00:21:57,400 --> 00:22:04,930
As I've noted, perceptions of police visibility are consistently linked with higher levels of trust and confidence in cross-sectional survey settings.

262
00:22:05,140 --> 00:22:09,130
I've always been real sceptical of vaccines, at least possible, at least as possible.

263
00:22:09,220 --> 00:22:14,470
The people who have relatively high levels of trust and confidence believe that they see the police enough or believe that

264
00:22:14,470 --> 00:22:20,740
they see the police quite often than I used to think that the causality in most undoubtedly is flowing in both directions.

265
00:22:21,100 --> 00:22:24,850
And underlying this, I think, is all of us have concerns. I can avoid that.

266
00:22:24,850 --> 00:22:28,419
Do people really notice police activity and their neighbours might have got more

267
00:22:28,420 --> 00:22:31,510
interesting and more important things to do than the current constantly noticing,

268
00:22:31,630 --> 00:22:38,170
noticing whether police officers are patrolling in their areas and of course that is visibility causally linked to trust.

269
00:22:38,350 --> 00:22:45,400
And this experiment is part of a much wider study into mounted police that Chris and I have been doing for the last 18 months.

270
00:22:45,490 --> 00:22:49,630
We think it's probably the first study into massive policing has ever taken place anywhere in the world.

271
00:22:49,780 --> 00:22:51,730
It is very niche, obviously,

272
00:22:52,330 --> 00:23:00,220
but I think it uncovers lots of interesting elements and interesting aspects of policing in a context such as England and Wales.

273
00:23:01,330 --> 00:23:06,910
So what did we do, which is a quasar experiment? It's a crazy experiment because there's no randomisation of the treatment.

274
00:23:07,270 --> 00:23:11,230
So what we did is we took six areas, four in Gloucestershire and two in London.

275
00:23:11,500 --> 00:23:16,990
The two are in Gloucester itself. Two are in the wider county of Gloucestershire and two were in south London.

276
00:23:17,290 --> 00:23:22,000
They were mostly wards. One was slightly bigger than an actual ward. We matched them basically on where they were.

277
00:23:22,150 --> 00:23:26,320
So we had the two areas in Gloucester, two areas in Gloucestershire, the two areas in south London.

278
00:23:26,500 --> 00:23:29,650
They were actually very similar to each other across a lot of different measures.

279
00:23:30,520 --> 00:23:34,899
And then we decided that three of those areas were going to get a dose of mounted policing.

280
00:23:34,900 --> 00:23:38,740
They were the test areas for those areas. We're not going to get a dose of match policing.

281
00:23:38,860 --> 00:23:45,370
Those were the control areas and we conducted first pre-test telephone survey of a sample size of a thousand.

282
00:23:45,370 --> 00:23:47,559
That was in February this year. Then in March,

283
00:23:47,560 --> 00:23:55,780
the test areas got the dose of mounted policing is basically was seven or eight six hour shifts at mass patrols in these areas over the course.

284
00:23:56,280 --> 00:24:02,859
So it wasn't very much. And as I said, it's showing the minute the actual public facing time with the mounted units was about two and a half hours,

285
00:24:02,860 --> 00:24:08,169
2 to 3 hours in the areas where they patrolled business as usual in the control sites.

286
00:24:08,170 --> 00:24:16,870
So basically foot patrols. And then alongside this, in both the tests and the control sites, we did a set of systematic social observation.

287
00:24:16,960 --> 00:24:21,040
Basically, we followed the horses around and we had a huge amount of help from students to do this.

288
00:24:21,040 --> 00:24:24,999
Some of them are in the room at the moment, so thanks very much. And literally, we followed Ramstein.

289
00:24:25,000 --> 00:24:31,360
It forces us for two and a half hours as they walked around these two neighbourhoods engaging will do whatever it was they were doing.

290
00:24:31,810 --> 00:24:34,510
Chris developed a mobile phone app to help us do this.

291
00:24:34,630 --> 00:24:38,800
So as we were following the forces around that were tapping into our mobile mobile phones, what was happening,

292
00:24:38,800 --> 00:24:44,860
how many people were talking to them, the quality of the interaction, etc., etc. And then this was automatically uploaded to a database.

293
00:24:44,860 --> 00:24:51,190
So the analysis was very, very useful, fantastic tool. And then we have hotels post test sample poster stuff as well.

294
00:24:51,460 --> 00:24:54,580
So I had a match pass pre post design as to the technical phrase.

295
00:24:55,090 --> 00:24:55,510
Well, that's it.

296
00:24:55,660 --> 00:25:01,900
Here's some of the things we see, some of the images, what they look like actually this I mean, I think, Richard, you took all of these, didn't you?

297
00:25:02,800 --> 00:25:08,260
This is relatively easy because usually as soon as they stop like this, as well as I can show, they were surrounded by people.

298
00:25:08,620 --> 00:25:09,910
So as soon as don't help match it,

299
00:25:10,090 --> 00:25:15,219
officers would stop in a shopping centre in the centre of town that act like a magnet to people who would literally,

300
00:25:15,220 --> 00:25:20,710
in some cases, swarm around them, patting horses, talking. Still, that's probably really, really important.

301
00:25:20,920 --> 00:25:24,730
But these were just. Neighbourhood patrols in the classic sense.

302
00:25:24,940 --> 00:25:28,690
They just literally wandered around the neighbourhoods that we told them to go wander around.

303
00:25:28,720 --> 00:25:30,520
They may have thought they were doing something different,

304
00:25:30,520 --> 00:25:39,370
but they really wanted three hypothesis first weapons used in actual experiments, tests that they kind of we need to prove those.

305
00:25:39,370 --> 00:25:41,650
We need to have some evidence relation says before we get to the third one.

306
00:25:41,800 --> 00:25:48,100
So the first hypothesis is where people would notice this mode of policing and as I say, is open to some question that people notice policing in that.

307
00:25:48,370 --> 00:25:53,440
And then April's second hypothesis that seeing mass patrols would be associated with high levels of trust and confidence.

308
00:25:53,710 --> 00:25:59,050
And in this study, we distinguish between these components of trust and confidence, fairness, effectiveness, community engagement.

309
00:25:59,140 --> 00:26:01,570
And we had overall measures of confidence as well.

310
00:26:01,720 --> 00:26:08,230
And the third hypothesis we tested with the experimental data content of the police would increase in test sites, rates of control sites.

311
00:26:09,670 --> 00:26:14,360
This is how we measured the components of trust. I'm sure you can't see those. But trust implies community engagement.

312
00:26:14,400 --> 00:26:16,930
And then this is this is probably the most important one in this context.

313
00:26:17,080 --> 00:26:20,180
So we measured that by questions that I understand to the police in this area.

314
00:26:20,200 --> 00:26:25,570
Do you agree or disagree? The police in this area understand issues that matter to people in the area in which you live,

315
00:26:25,870 --> 00:26:31,270
engage with all members of the public in which you live, attacking the issues that matter to people in the area.

316
00:26:31,510 --> 00:26:35,500
Shoot. If you see what we're trying to get to with community engagement and trust in police force,

317
00:26:35,680 --> 00:26:39,880
do they keep you treat people with dignity and respect on a friendly and approachable effectiveness?

318
00:26:40,120 --> 00:26:44,470
Are they effective at preventing crimes? Are the effects of catching people who commit crimes?

319
00:26:44,710 --> 00:26:49,150
And overall confidence. And that's got a measure of whether you would report a crime in the future.

320
00:26:49,810 --> 00:26:53,890
How confident are you in the ability of the police to deal with crime and disorder issues in your neighbourhood?

321
00:26:54,100 --> 00:26:57,250
And taking everything to count how good a job developers do.

322
00:26:57,460 --> 00:27:01,590
Overall measure of confidence. First of assisted people.

323
00:27:01,600 --> 00:27:05,290
Notice the match patrols. Absolutely. I'm stunned by this data.

324
00:27:05,290 --> 00:27:13,510
Every time I look at it in the test areas, the proportion of people reporting that had recently seen massive police rise from 15% to 43%.

325
00:27:13,660 --> 00:27:18,910
Now, I should say, none of these areas have had regular amounts of patrols before in living memory.

326
00:27:19,510 --> 00:27:22,660
The 15% in the three test periods comes all from London.

327
00:27:22,900 --> 00:27:26,380
So people in London are just more used to seeing horses in some general sense.

328
00:27:26,500 --> 00:27:29,730
So they were saying, yes, I had seen massive patrols in the neighbourhoods. They hadn't.

329
00:27:29,740 --> 00:27:34,209
There's not a typical. This is the kind of stuff you get with survey data in one area.

330
00:27:34,210 --> 00:27:38,110
This is our ancestor in Gloucestershire of roughly half the town,

331
00:27:38,230 --> 00:27:42,430
according to this data noticed that had been in recent months of patrols in their area.

332
00:27:43,150 --> 00:27:47,110
It's a massive place associated with high levels of trust and confidence. Again, definitely.

333
00:27:47,350 --> 00:27:51,280
So 79% of those who had recently seen mass patrols agreed to the statement.

334
00:27:51,430 --> 00:27:54,790
The police understand the issues that matter to people in the area in which you live,

335
00:27:55,030 --> 00:27:58,270
compared to 69% of those who hadn't recently seen police control.

336
00:27:58,420 --> 00:28:00,850
And I find again, I find this stunning. Let's think about that for a second.

337
00:28:01,060 --> 00:28:07,270
Merely having seen an officer on horseback is associated with ten percentage point shift in people, agreeing.

338
00:28:07,270 --> 00:28:11,260
The police understand the issues, the matter of people in the area in which they live.

339
00:28:12,010 --> 00:28:15,819
Make of that what you will purposes.

340
00:28:15,820 --> 00:28:17,410
The analysis is differences and differences.

341
00:28:17,410 --> 00:28:24,430
So what we're doing is looking at change in the test sites relative to change in the control sites because get pre and post measures in both.

342
00:28:24,670 --> 00:28:27,460
So this chart plots the differences in differences coefficients.

343
00:28:27,730 --> 00:28:35,650
If the if if the dot is above the zero line, that means there was a positive change in the test sites compared with the control sites.

344
00:28:35,920 --> 00:28:44,350
So all those dots mean three of those dots are positive. But error bars indicate there was two cities tested statistically divisible from zero.

345
00:28:44,650 --> 00:28:48,430
But because we've done the system systematic social observation,

346
00:28:48,610 --> 00:28:55,330
we knew that the patrols in Gloucester had been going outside the area in which they were meant to be patrolling into the city centre,

347
00:28:55,510 --> 00:29:00,820
where they all must undoubtedly have been seen by people living all over Gloucester, including those living in the control sites.

348
00:29:01,000 --> 00:29:07,120
So we had treatment, migration at people in the control site experience the treatment which is meant to be limited to the test sites.

349
00:29:07,360 --> 00:29:12,700
When we looked in the data, yes, the proportion of people in the control site in Gloucester saying that they'd recently seen matched

350
00:29:12,700 --> 00:29:17,559
security patrols in the neighbours had increased significantly and they hadn't recently seen mass patrols.

351
00:29:17,560 --> 00:29:23,230
And that was because I wanted it. I seen them in the city centre, but you get the idea this is the way that people answer certain questions.

352
00:29:23,560 --> 00:29:27,540
So when we take Gloucester out, which I think we probably should be, I think this best,

353
00:29:27,550 --> 00:29:32,260
these results should be considered together because it's not entirely clear whether one is right and one is wrong.

354
00:29:32,440 --> 00:29:39,640
But when we take Gloucester out, we find that the map of patrols has a significant and positive effect on trust in police community engagement,

355
00:29:39,850 --> 00:29:44,080
trust in police fairness and overall confidence, and almost on effectiveness as well.

356
00:29:44,410 --> 00:29:50,500
All those coefficients are positive and actually quite large. It's it's not quite statistically significant and conventional.

357
00:29:53,450 --> 00:29:57,800
I would say the systematic social innovation was integral to this experiment.

358
00:29:57,840 --> 00:30:05,360
It never would have worked as well as it did or at all if we hadn't have been doing systematic social observation at the same time.

359
00:30:05,720 --> 00:30:07,520
So we sat out together collectively.

360
00:30:07,520 --> 00:30:13,280
That was, as I say, that was a lot of this quite a large number of us involved in this increase in people from place to place in Gloucestershire.

361
00:30:13,430 --> 00:30:21,350
We set out to take shifts in both the test and control sites, which is 64 hours of public facing police time in total.

362
00:30:21,770 --> 00:30:26,450
We recorded over 5600 what we termed engagements between police and public.

363
00:30:26,630 --> 00:30:31,740
What we were interested in doing was trying to notes, trying to record people, noticing police.

364
00:30:32,300 --> 00:30:33,350
That's obviously quite difficult.

365
00:30:33,680 --> 00:30:39,860
I mean, anyone, for instance, sees a police officer when they're walking down the street, even if it's only to step out of their way.

366
00:30:40,040 --> 00:30:43,970
We were looking for something slightly more than that, some kind of active recognition.

367
00:30:44,300 --> 00:30:49,730
So what we tended to do was record. What we did do is record things like if someone stopped and pointed at the horse.

368
00:30:50,150 --> 00:30:53,690
If someone took a camera, it's a photo of the horse, of the officer.

369
00:30:53,960 --> 00:30:58,040
If someone punches their child and we could see that their child was registered those because those

370
00:30:58,040 --> 00:31:02,420
are the things that we think people might remember when stimulated to set by the survey instrument.

371
00:31:02,630 --> 00:31:08,990
Further down the line, we also recorded more in-depth interactions between officers and members of the public.

372
00:31:09,200 --> 00:31:17,180
And what we found is the horses generated six times as many as many of these basic encounters as the four officers in the control size.

373
00:31:17,330 --> 00:31:23,780
This is really getting to understand just how noticeable these things are to people when they're walking around in neighbourhood settings.

374
00:31:24,470 --> 00:31:32,240
There wasn't that many of a statistically significant difference between the officers in the control sites and the match officers in the test sites.

375
00:31:32,420 --> 00:31:38,040
But what we did find is when it came to slightly longer, more in-depth interactions with officers,

376
00:31:38,090 --> 00:31:42,650
that people actually talking to them more than say hello when there was a horse there.

377
00:31:42,710 --> 00:31:44,690
These encounters tends to be more positive.

378
00:31:45,140 --> 00:31:50,420
So it's something about the force that seems to be enabling positive encounters between police and members of the public.

379
00:31:50,700 --> 00:31:56,660
But the same number roughly translates to neutral sites. But those in the test sites were more positive.

380
00:31:58,250 --> 00:32:06,080
So why did it work? Which is what is I mean, I think the experiment works because the intervention was clear, it was precise, it was simple.

381
00:32:06,200 --> 00:32:09,170
It was very, very different from business as usual.

382
00:32:09,260 --> 00:32:16,160
I find the system a6's observation greatly aided in its acts and results, massively aided interpreting the results.

383
00:32:17,450 --> 00:32:21,079
It is the ultimate cause of the observed effects.

384
00:32:21,080 --> 00:32:23,960
This is where it starts. Get a bit interesting. We're not sure.

385
00:32:24,410 --> 00:32:30,890
We don't know whether it was the horse, the rider or the horse and the rider and also their implications for that.

386
00:32:31,010 --> 00:32:37,190
And one of the interesting things that we can never know that because you can't separate out the horse and the rider in this context, it's impossible.

387
00:32:37,310 --> 00:32:43,730
So this intervention is something I'll come back to towards the end of his intervention, although it seems quite simple, in fact, it is quite simple.

388
00:32:43,820 --> 00:32:47,809
It's also quite complicated. We don't know what it is, what combination of horse, right.

389
00:32:47,810 --> 00:32:50,840
And horse and what caused the effects that we observed.

390
00:32:51,530 --> 00:32:52,129
What are the lessons?

391
00:32:52,130 --> 00:32:58,130
Well, some of the lessons here are quite clear, and I think police deployments in neighbourhood patrol can increase public trust.

392
00:32:58,820 --> 00:33:02,270
And this probably has something to do with both visibility and engagement.

393
00:33:02,390 --> 00:33:06,560
I think it's probably something to do with both being there and talking to people and

394
00:33:06,560 --> 00:33:10,400
some combination of those two things that lets the increase in public trust encompass,

395
00:33:10,550 --> 00:33:15,350
hence the community engagement measures and the large effects that the mass patrol seemed to have on this.

396
00:33:15,560 --> 00:33:19,480
I mean, our lessons here are much less clear. I mean, do we put all neighbourhood police on horses?

397
00:33:19,490 --> 00:33:26,000
Is that the lesson here? Well, clearly not. So what the policy lessons here for police managers, how long do these effects last?

398
00:33:26,240 --> 00:33:30,680
We have designed that the experiment has always an issue of any kind of intervention like this.

399
00:33:30,860 --> 00:33:36,140
How long did it take? People's views in the test sites to regress back to their previous level?

400
00:33:36,500 --> 00:33:41,690
I would suggest probably not at all. Ashley And another issue here, this is in a sense, a much bigger issue.

401
00:33:41,750 --> 00:33:45,950
One lesson from this and actually the way it's being interpreted is improving some sections of police service.

402
00:33:46,550 --> 00:33:52,310
Is that all you have to do if you've got trouble just in case confidence in particular is put more police officers in there.

403
00:33:52,760 --> 00:33:57,560
And that in and of itself will wave a magic wand and and solve the problem that you have.

404
00:33:58,070 --> 00:34:02,270
That's extremely unlikely to be the case. And it also raises the issue of how much policing is too much.

405
00:34:02,570 --> 00:34:06,470
Is there a sweet spot here where you give people some extra police and it's fine,

406
00:34:06,650 --> 00:34:11,090
you give them a lot more and it starts to be seen as an occupation at the outset.

407
00:34:11,090 --> 00:34:14,840
But it's worth asking the question. Another issue and again, we'll return to this at the end.

408
00:34:14,960 --> 00:34:22,700
Does everyone react to visible policing in the same way? Highly visible policing in the same way as they seem to have done in this experiment.

409
00:34:22,820 --> 00:34:29,030
Again, surely not just one experiment which worked well as an experiment.

410
00:34:29,030 --> 00:34:35,500
It worked extremely well as an experiment. The second study I want to talk about Scots, that study, what's called that.

411
00:34:35,510 --> 00:34:43,430
I mean, it actually worked much less well as an experiment as I would go on to explain, but is nonetheless so extremely interesting.

412
00:34:43,850 --> 00:34:47,989
This really harks back to right to the core of the procedural justice model.

413
00:34:47,990 --> 00:34:51,650
That's the motivation for this study. So we're moving away from visibility.

414
00:34:51,930 --> 00:34:56,089
Per say I perhaps a bit of chatting between police and public security,

415
00:34:56,090 --> 00:35:00,620
trying to get the police to operationalise some of the core elements of the procedural

416
00:35:00,620 --> 00:35:04,580
justice model in the way they're dealing in the way that they're talking with people.

417
00:35:04,790 --> 00:35:12,110
It's got to that was at its inception was intended to be a replication of the Queensland Community engagement trial.

418
00:35:12,200 --> 00:35:20,690
Q So is a study done by Ray Massaro and her team and I think 2010 in Queensland as you is, you'd expect in Australia.

419
00:35:20,930 --> 00:35:28,100
And this was looking at the effects of a procedural justice intervention in the context of mass random breath testing in Queensland.

420
00:35:28,400 --> 00:35:33,410
So I don't know if you've been Australia, you've seen this and what they do in Australia, they do it in lots of states.

421
00:35:33,410 --> 00:35:39,300
I think they set up huge roadblocks and they put in large numbers of people on a more or less random basis.

422
00:35:39,570 --> 00:35:45,170
Obviously it's always unclear what random means in that context, but it seems random to the drivers, I suspect,

423
00:35:45,380 --> 00:35:52,220
and the baseline encounter in set where they built their intervention on seemed to be something like the officer would march up to the car,

424
00:35:52,460 --> 00:35:58,790
he'd he or she would thrust the breathalyser into the window, blow in this, the person blowing this, the reading would come back negative.

425
00:35:58,910 --> 00:36:03,110
They would say on your way. They averaging counselling from the baseline in case that was 30 seconds.

426
00:36:03,740 --> 00:36:09,590
That's important. So what they did in that was design an intervention that was intended to increase the pursuit

427
00:36:09,590 --> 00:36:14,300
of justice of that encounter from the perspective of the members of the public being stopped.

428
00:36:14,510 --> 00:36:19,340
So they started off with just getting the officers to say hello. My name is Officer 75 from such and such station.

429
00:36:19,490 --> 00:36:25,060
Today we're doing random breath tests. This is why we're doing it. They hadn't been doing that before, and then they felt more and more surprising,

430
00:36:25,110 --> 00:36:32,210
just as modelled on the intervention and the average encounter length in the in the test area in Q was one was 90 seconds.

431
00:36:32,630 --> 00:36:35,960
So they kind of tripled the length of the encounter, but it was still only 90 seconds long.

432
00:36:36,170 --> 00:36:41,090
And they found this intervention had a positive effect on perceptions of police fairness during the encounter,

433
00:36:41,360 --> 00:36:49,370
satisfaction with the encounter overall, and trust in police legitimacy in a more general sense and case that we aim to replicate this in

434
00:36:49,370 --> 00:36:54,049
the context of race policing in Scotland there is a reputation of a randomised controlled trial,

435
00:36:54,050 --> 00:37:01,190
treatment of the experiment and we were examining a high volume routine encounters between members of the public and police officers.

436
00:37:02,510 --> 00:37:06,469
However, when we started looking at what was going on in Scotland,

437
00:37:06,470 --> 00:37:09,710
it was very apparent to us that it was very different to what was going on in Scotland.

438
00:37:10,010 --> 00:37:13,370
So the most obvious place to implement this trial was around what they call their

439
00:37:13,370 --> 00:37:17,410
national festive road safety campaign that they run every year over Christmas for very,

440
00:37:17,480 --> 00:37:22,990
very, very obvious reasons. During this campaign, drivers stopped with the aim of preventing drunk driving.

441
00:37:23,210 --> 00:37:28,070
Although in Scotland they're not allowed to do random breath tests, they have a reason to breathalysed someone.

442
00:37:28,280 --> 00:37:32,480
And for improving vehicle and driver safety on the graves during the Scottish winter.

443
00:37:32,960 --> 00:37:39,300
At the beginning, Police Scotland estimates have been around 20,000 encounters doing this right road safety festival.

444
00:37:39,320 --> 00:37:44,020
If it's called race but never say the festive road safety campaign says high volume.

445
00:37:44,030 --> 00:37:49,249
There's a lot of contact between police and public in this campaign. And again, we have match pairs, pre posters.

446
00:37:49,250 --> 00:37:56,270
That's why the song was actually different excuse as well. So there are 20 race police units in Scotland's be divided into ten pairs

447
00:37:56,780 --> 00:38:00,560
primarily on geographical factors were some other factors were important as well.

448
00:38:00,800 --> 00:38:06,620
Then we randomly assigned one of those pairs to the test group and then the other one went into the control group.

449
00:38:07,370 --> 00:38:15,740
All drivers stopped over the last period in Scotland were given a questionnaire to some of them it says some of them were in control.

450
00:38:17,740 --> 00:38:22,610
But our main issue was a business in usual. Scotland was different to that in Australia.

451
00:38:22,670 --> 00:38:26,990
Scottish police were already doing a lot of things that the perceived justice model suggests they shouldn't do.

452
00:38:27,290 --> 00:38:31,939
So they were much better at talking straight, so much better explaining to Travis what had been stuck there,

453
00:38:31,940 --> 00:38:34,610
much more in some sense, friendly with Travis.

454
00:38:34,610 --> 00:38:41,840
I mean, I think you want to put some caveats around that, but it was clearly a very different kind of encounter than what was going on in Australia.

455
00:38:42,050 --> 00:38:47,270
So we had, we were faced with the challenge of how can we design an intervention and how can we design the intervention to

456
00:38:47,270 --> 00:38:53,630
make our intervention as different from business as usual as possible within the context of these encounters.

457
00:38:53,900 --> 00:38:58,220
So what we did is we designed a set of working with the police, with the focus groups and what have you around this.

458
00:38:58,730 --> 00:39:01,969
We designed a kind of checklist or move in that really boiled down.

459
00:39:01,970 --> 00:39:08,180
It's a checklist that attempted to show that they would verbally communicate all of a series of key messages

460
00:39:08,300 --> 00:39:14,450
based around the pursuit of justice model during their encounters with people during the road safety campaign.

461
00:39:14,600 --> 00:39:16,370
We also added a leaflet to this,

462
00:39:16,370 --> 00:39:23,749
so all the drivers in the test condition got a leaflet so that the intervention was both making sure the officers did all their things

463
00:39:23,750 --> 00:39:29,479
on that checklist in terms of procedural justice and gave people a leaflet which was intended to explain people why they'd been stopped.

464
00:39:29,480 --> 00:39:35,000
The importance of the campaign. So. Rankin Communication of why police are doing this.

465
00:39:35,180 --> 00:39:41,660
So the key message was obviously respect, equality, trustworthy motives, dignity, they try to seize it and passive participation,

466
00:39:41,750 --> 00:39:47,870
etc., etc. All the core elements of social justice were meant to go into these experimental encounters.

467
00:39:48,620 --> 00:39:51,590
And that's the checklist. Again, you can't say that, but just in case anyone is interested.

468
00:39:51,770 --> 00:39:55,500
The question is this his offices were given an 858 memoir to carry around,

469
00:39:55,580 --> 00:39:59,090
and they were supposed to remember to do all these things during the encounter.

470
00:39:59,720 --> 00:40:05,480
That's the leaflets got a bit distorted when I copied it out, but basically it starts off, for example.

471
00:40:05,600 --> 00:40:09,650
Thank you for your time today. We welcome your views on all aspects of police in Scotland.

472
00:40:09,860 --> 00:40:13,310
Details of our latest initiatives local police and teams as contest. We found a website.

473
00:40:13,430 --> 00:40:16,670
I wanted a phone number on there, but they wouldn't do that. They wouldn't give us a phone.

474
00:40:17,480 --> 00:40:21,210
But you get the idea for hypothesis.

475
00:40:21,350 --> 00:40:25,040
Most of us in this experiment, we had an intervention.

476
00:40:26,060 --> 00:40:33,290
They just explained first hypothesis they would improve perception of perceived justice and trust with the officers doing the stop itself.

477
00:40:33,830 --> 00:40:41,320
Second hypothesis it would increase overall satisfaction with the overall satisfaction why the officers behave doing and stop further officers.

478
00:40:41,330 --> 00:40:45,020
It would increase trust and confidence in the police in Scotland in a general sense.

479
00:40:45,200 --> 00:40:49,639
So we got people's perceptions of specific officers conduct in the stock and their perceptions of

480
00:40:49,640 --> 00:40:54,400
the police in Scotland more widely and for psychosis that the intervention would increase police,

481
00:40:54,410 --> 00:40:58,310
but just the perception of police legitimacy in Scotland as a whole.

482
00:40:58,790 --> 00:41:04,760
And just as I promised, I'd come back to this and this is how we measure legitimacy in this study and in lots of studies as well.

483
00:41:04,880 --> 00:41:10,790
So we distinguish between different two different components of legitimacy, what we can call perceived duty to it by the police.

484
00:41:10,940 --> 00:41:17,239
I feel a moral obligation to pay the police. I feel a moral duty to support the decisions of police officers, even when I disagree with them.

485
00:41:17,240 --> 00:41:22,340
I think you're trying to have you see what we're trying to get out there and what we call moral or normative alignment with the police.

486
00:41:22,520 --> 00:41:27,380
The police have the same sense of right or wrong as me. Please stand up for values that are important for people like me.

487
00:41:27,620 --> 00:41:31,849
I support the way police usually act in our systems.

488
00:41:31,850 --> 00:41:35,650
Again, differences and differences. So we were looking at how we have pre post-match press.

489
00:41:35,670 --> 00:41:38,810
We had a survey in the field in the first week of December.

490
00:41:39,170 --> 00:41:45,020
We then took a week off when we distributed the survey questionnaires and all the materials to the officers and the experimental units.

491
00:41:45,200 --> 00:41:50,690
And then we had a survey running from that first, second week right to the end of the paper, which I think was the 2nd of January.

492
00:41:50,840 --> 00:41:54,469
So pre and post measures again. So again, we're looking at differences in differences.

493
00:41:54,470 --> 00:42:01,440
So the analysis is exactly the same as before. Why is the change in the test size compared with the change in the control sites?

494
00:42:01,640 --> 00:42:08,750
Again, a positive coefficient would indicate this positive change in the test sites compared with the control sites or the coefficients and negative.

495
00:42:09,350 --> 00:42:13,040
We didn't make things better. We didn't even look at effect. We made things worse.

496
00:42:13,640 --> 00:42:22,160
So in the control sites, perceptions of stop procedure justice fell compared to the control sites just doing the also found that it wasn't quite

497
00:42:22,160 --> 00:42:28,400
statistically significant and satisfaction with the SOC significantly in the control sites compared with the test sites.

498
00:42:28,970 --> 00:42:33,080
I can't say this again enough. We didn't make things better. We made things worse.

499
00:42:34,850 --> 00:42:37,670
Similarly, with overall trust, I mean, they're not statistically significant.

500
00:42:37,740 --> 00:42:41,570
Again, they're both never negative and that's actually quite large as these things go.

501
00:42:42,020 --> 00:42:46,880
And similarly, illegitimacy, again, negative, not statistically significant, nothing said.

502
00:42:48,160 --> 00:42:51,410
So I prefer this one. We have the data didn't support our forces.

503
00:42:51,410 --> 00:42:59,750
In fact, the data contradicted intervention, damaged perceptions of severe justice to in of intervention decreased satisfaction of the stock.

504
00:43:00,230 --> 00:43:07,640
There's no significant effect on general trust and confidence. Nice checks on legitimacy for all the statistical effects in the study were negative.

505
00:43:07,760 --> 00:43:12,590
So in some overall sense I'm quite confident saying we just made things worse.

506
00:43:13,010 --> 00:43:16,160
Why is that? Why didn't it work in this context?

507
00:43:16,400 --> 00:43:22,520
Well, I think there's a number of reasons. For one thing, again, it was important to notice again, it seems quite simple.

508
00:43:22,520 --> 00:43:25,070
We've got a really quite complicated intervention here.

509
00:43:25,370 --> 00:43:30,980
Not only have we got two aspects of it, we've got the checklist and things the officer wants to do and the leaflet,

510
00:43:31,160 --> 00:43:36,650
but we've got all the different things that the police officers were meant to do. We didn't know which one of those things they were meant to do.

511
00:43:36,830 --> 00:43:40,280
We had the effect that we experience in which combination of those things.

512
00:43:40,610 --> 00:43:43,970
I think what really happened is we bureaucratised the encounter.

513
00:43:44,420 --> 00:43:51,050
So to the extent that the officers did the things that we told them to do, and of course that's always open to some doubt in this context.

514
00:43:51,650 --> 00:43:54,950
I think they become more became more worried about ticking boxes.

515
00:43:55,160 --> 00:43:57,770
They became more worried about doing all the things that are meant to do.

516
00:43:57,950 --> 00:44:01,910
And in doing so, they became less interested in the person they were dealing with.

517
00:44:02,210 --> 00:44:10,850
I also suspect the process took longer in the experimental sites and this was precisely experienced is precisely unfair by the people involved.

518
00:44:11,030 --> 00:44:14,720
Because if you take if you drag something out, if something takes longer than it really needs to,

519
00:44:14,810 --> 00:44:18,110
I think you're really communicating disrespect to someone by inside.

520
00:44:18,110 --> 00:44:22,339
I don't care about the time. I'm going to keep you here as long as I need to. And there's some evidence from that.

521
00:44:22,340 --> 00:44:25,700
And Q said, because in Q set a time the length of the encounters.

522
00:44:25,970 --> 00:44:31,820
So the main encapsulated from Q set was 90 seconds but had a positive effect on trust, legitimacy,

523
00:44:31,820 --> 00:44:37,940
etc., etc. When the encounter length went, I think it's over something like 130 seconds.

524
00:44:38,090 --> 00:44:42,980
That effect reversed and then the intervention had a negative effect on trust, which is messy cooperation.

525
00:44:43,250 --> 00:44:47,060
That's probably something. It could be the leaflet. I'm actually don't think so.

526
00:44:47,360 --> 00:44:51,290
Various raised them 20 people in the experimental group couldn't.

527
00:44:51,360 --> 00:44:55,080
Remember, Ganguly fell over, couldn't remember, didn't gain the leaf, didn't get the leaflet.

528
00:44:55,230 --> 00:44:58,350
Their views were much, much more negative, and everyone else in the experiment took it.

529
00:44:58,530 --> 00:45:00,600
So I don't think it was any favour. Of course it could have been.

530
00:45:00,960 --> 00:45:05,430
And there's a follow up study in the pipeline to really try and get to grips with some of the ideas.

531
00:45:05,580 --> 00:45:10,530
So, again, we're conducting focus groups with the cops involved in the intervention to see,

532
00:45:10,530 --> 00:45:14,430
you know, what went wrong, how could we do this better in the future?

533
00:45:15,000 --> 00:45:18,060
So I'm coming to the end now, has a few final slides.

534
00:45:18,210 --> 00:45:23,070
First, reflecting on experiments of experiments in public companies, places you see,

535
00:45:23,220 --> 00:45:27,600
and I think possibly in criminology and criminological settings more broadly.

536
00:45:27,830 --> 00:45:31,469
And I think one important message here, and this should be fairly obvious,

537
00:45:31,470 --> 00:45:37,049
but I think it's important to reiterate experiments that really work in isolation and even much,

538
00:45:37,050 --> 00:45:43,230
much more effective, most effective, properly effective when they're embedded with a very set of a modes of research,

539
00:45:43,230 --> 00:45:46,980
as we found in the matter of community patrols, is systematic.

540
00:45:46,980 --> 00:45:49,950
Socialisation was a vital aspect of that piece of research.

541
00:45:50,220 --> 00:45:55,920
And of course, in terms of replication, replication is a powerful device, but as we saw, replication can be difficult.

542
00:45:56,220 --> 00:46:00,150
So one of the one of the implications, I think from an example publishing, Scott said,

543
00:46:00,360 --> 00:46:04,110
and I suspect it will be treated by most people as a replication of CUSA.

544
00:46:04,230 --> 00:46:07,910
It really wasn't. It was another type of experiment altogether.

545
00:46:07,920 --> 00:46:10,920
But we treat it as a replication that has implications.

546
00:46:11,670 --> 00:46:14,480
This touches on a point I raised before. We've considered average effects.

547
00:46:14,490 --> 00:46:18,090
All statistical analysis at the top of your talking right is looking at average effects.

548
00:46:18,270 --> 00:46:19,440
So on average,

549
00:46:19,770 --> 00:46:26,960
people in the communities who experienced the massive patrols had higher levels of trust and confidence as a result of those patrols, we think.

550
00:46:26,970 --> 00:46:31,080
But that doesn't mean to say that everyone in those communities felt the same way about those mounted patrols.

551
00:46:31,200 --> 00:46:36,029
It's possibly quite likely that some people came away with more negative views of the police.

552
00:46:36,030 --> 00:46:39,450
As a result, have views were just swamped by the positive results of people?

553
00:46:39,570 --> 00:46:45,240
And how to isolate this variation of in different population groups is a real challenge for both this type of work and,

554
00:46:45,240 --> 00:46:49,290
as I say, experiments in criminological ecological settings more widely.

555
00:46:49,440 --> 00:46:55,259
And things can go right, things go wrong as well as rates, or more likely no effects, to be honest.

556
00:46:55,260 --> 00:46:58,570
But things can get really quite seriously. Well, what do you do with that?

557
00:46:58,590 --> 00:47:04,469
What do you do with that information? I think that has best theoretical and policy implications in a kind of philosophy of science

558
00:47:04,470 --> 00:47:09,780
sense in Scots that we designed an intervention based on the principles of pursuit of justice.

559
00:47:10,020 --> 00:47:13,920
We spent quite a lot of time doing that. We've built all those principles into that.

560
00:47:14,070 --> 00:47:17,820
We really tried our best to make this work in this context. We made things worse.

561
00:47:18,060 --> 00:47:21,720
So we should go back and revisit your theory because there's something wrong with our theory. It's not doing.

562
00:47:21,840 --> 00:47:26,010
It's not explaining the world in the way we think it should do. That's not going to happen.

563
00:47:26,010 --> 00:47:29,219
I can guarantee you. No one is going to go back to procedures just tomorrow, I think.

564
00:47:29,220 --> 00:47:35,010
Well, there's something wrong with this theory of back of Scots now. And I think in the case of justice, I think that's justified.

565
00:47:35,160 --> 00:47:39,270
There's lots and lots and lots of other evidence out there suggest the pursuit of suspects are real,

566
00:47:39,420 --> 00:47:44,880
not do persist in the world in some sense, obviously in other contexts that might not be the case.

567
00:47:45,450 --> 00:47:51,929
And there probably I think there's always a temptation in experimental research to think that you've got the intervention wrong and

568
00:47:51,930 --> 00:47:57,210
all you need to do is redesign the intervention rather than revisiting the theory that motivates the intervention in the first place.

569
00:47:57,390 --> 00:48:02,320
In policy terms, do we want police in Scotland to come away from this saying not only no,

570
00:48:02,340 --> 00:48:06,240
we don't need to worry about pursuing justice, as our officers are already doing quite well at this.

571
00:48:06,510 --> 00:48:09,870
But well, anything we try to do about procedures, justice isn't going to make things better.

572
00:48:09,870 --> 00:48:15,719
It's going to find things worse. What obviously not. And of police Scotland police in Scotland off which fortunately not that stupid but

573
00:48:15,720 --> 00:48:20,370
again I think you can see the risks that may make may come out of some of this stuff.

574
00:48:20,630 --> 00:48:26,310
I mean, I think much more broadly, of course, experiments can create as many questions as science is the answer.

575
00:48:26,310 --> 00:48:30,930
And and complex interventions are particularly problematic in this.

576
00:48:30,930 --> 00:48:34,650
But many of the interventions and lots of the experiments that people are doing with in

577
00:48:34,950 --> 00:48:40,439
criminology and a far more complicated than the interventions I've just described to you today.

578
00:48:40,440 --> 00:48:47,309
And even in this relatively simple context, we've seen the implications of trying to unpick what aspects of the massive trial had an effect

579
00:48:47,310 --> 00:48:52,800
on trust and what aspects of our intervention in Scotland that led to these negative outcomes.

580
00:48:52,890 --> 00:48:59,270
And it's obviously all speaks to the much more important idea. And yes, I was impressed by Nancy Cartwright, some of the people she's worked out.

581
00:48:59,430 --> 00:49:05,280
It's a problem for policymakers in particular of taking the results of randomised controlled trials and other types of experiments and saying,

582
00:49:05,280 --> 00:49:07,770
well, it works over here, it works in this context.

583
00:49:07,920 --> 00:49:13,350
Will it work in the context in which I'm operating in and complex interventions are make that really,

584
00:49:13,350 --> 00:49:19,200
really difficult to unpick in terms of kind of trust and confidence and legitimacy research.

585
00:49:19,200 --> 00:49:22,770
I think I think some of the key lessons Sarah and other people might have other ideas

586
00:49:23,010 --> 00:49:27,710
and it seems pretty obvious to me now and I've always been very sceptical about this,

587
00:49:27,720 --> 00:49:32,250
that the long term decline in public confidence I was talking about at the beginning

588
00:49:32,310 --> 00:49:36,540
probably really has got something to do with declining police visibility over that time.

589
00:49:36,840 --> 00:49:43,260
There's this decline in trust and confidence happened over a period when, for example, rural police stations were closed.

590
00:49:43,440 --> 00:49:50,610
The police basically withdrew from many rural areas of England and Wales when police officers were moving off the streets and its cars.

591
00:49:50,610 --> 00:49:56,309
And so I think. It's in different ways on that basis. And I think this is the kind of research we've done with the horses,

592
00:49:56,310 --> 00:50:02,010
it fits with some other work is restarting and says there's just a link between police visibility and trust and confidence.

593
00:50:02,460 --> 00:50:10,080
The question then becomes, what do you do that due to that? I mean, take one of the test areas in the trunk study.

594
00:50:10,080 --> 00:50:11,130
This is science tester.

595
00:50:11,400 --> 00:50:18,150
I'm pretty confident that people in science tester are getting about the right level of pacing they need in some objective sense.

596
00:50:18,300 --> 00:50:22,290
I don't think it's a very high crime area. The police are there. They do do foot patrols in the area.

597
00:50:22,440 --> 00:50:26,370
There is a police station and someone says that they will come if you call them, etc., etc.

598
00:50:26,580 --> 00:50:32,970
But the fact that they react so well. So these these massive patrols really indicate to me that they're not getting the kind of police that they want.

599
00:50:33,210 --> 00:50:39,630
So how do you offset their subjective needs for highly visible forms of policing that may not be particularly effective in reducing crime,

600
00:50:39,840 --> 00:50:46,230
doing disorder, etc., etc., but satisfy the needs and wishes of the communities you're meant to be serving as police officers.

601
00:50:47,700 --> 00:50:52,559
The results from Scott set really suggest to me that when you've got contacts with

602
00:50:52,560 --> 00:50:56,790
relatively high levels of trust and confidence across the population as a whole,

603
00:50:56,910 --> 00:50:59,340
which is what you've got in Scotland, which is what you've got in England, well,

604
00:50:59,520 --> 00:51:05,700
for all a significant variation within that population, it seems unlikely that I think actually to me reflecting on this,

605
00:51:05,880 --> 00:51:12,870
the public perceptions are unlikely to be boosted any further from their already relatively high levels by just a simple additive effect.

606
00:51:12,990 --> 00:51:16,440
We just need to do a bit more procedural justice and we're going to get returns to that.

607
00:51:16,440 --> 00:51:21,059
In terms of trust and confidence, legitimacy, we may need more radical interventions.

608
00:51:21,060 --> 00:51:25,530
I think which parts in particular target those groups in the population that don't have high levels of trust?

609
00:51:25,860 --> 00:51:30,210
We don't have that sort of dismissing. And of course, there's a whole other set of questions of how do you identify those groups?

610
00:51:30,390 --> 00:51:31,980
What kind of interventions can you do that?

611
00:51:32,100 --> 00:51:37,860
Do they do do they look like procedural justice interventions that the general population might respond well to or not?

612
00:51:37,950 --> 00:51:44,160
Or do they look sorry? Not something very quite really quite different. And I think two points before I finish.

613
00:51:45,750 --> 00:51:49,680
One way to look, almost everything I've said to say is really this is just took an operation,

614
00:51:49,680 --> 00:51:53,300
operations in the operation operationalisation of the kind of bleedin obvious right.

615
00:51:53,430 --> 00:51:58,800
Of course people want perceived policing. Of course, people want police officers to treat them fairly.

616
00:51:58,980 --> 00:52:01,740
Of course, people want physical signs of order in the community.

617
00:52:01,860 --> 00:52:07,710
We didn't need experiments and expending tens of thousands of pounds on these things to prove that was the case.

618
00:52:08,460 --> 00:52:11,550
But that's not how much we had around that. But I think the research is interesting nonetheless.

619
00:52:11,820 --> 00:52:17,490
But if that being the case, if it's so obvious that these are things the police should be do, why do they get it wrong?

620
00:52:17,490 --> 00:52:19,830
So break is relatively often.

621
00:52:20,070 --> 00:52:25,350
What are the things motivating the police not to behave in the ways I've been talking about today, which everyone kind of thinks they should.

622
00:52:25,560 --> 00:52:29,170
They think they should in those circumstances, but they don't always behave like that.

623
00:52:29,190 --> 00:52:33,510
Why is that? Why do police officers misbehave so relatively often?

624
00:52:33,660 --> 00:52:36,000
There's a whole other set of interesting questions asked.

625
00:52:36,000 --> 00:52:41,160
Yes, I think this should be a key direction kind of trust, confidence, legitimacy, research in the future.

626
00:52:41,280 --> 00:52:45,630
Turn the lens back onto the police service and think about the ways that motivate their behaviours.

627
00:52:46,260 --> 00:52:50,370
And the final point, and I'm still not quite sure I'm kind of phrasing this right,

628
00:52:50,610 --> 00:52:56,309
but one of my concerns here and again this is a concern it's both specific in a trust is we see

629
00:52:56,310 --> 00:53:01,440
research and I think applies in other areas of criminal justice research as well is there's a risk of

630
00:53:02,130 --> 00:53:07,230
what I can only characterise as insurance lies in the aims of criminal justice research and seems to

631
00:53:07,230 --> 00:53:13,410
me the pursuit of justice policing process based policing models are good in themselves and needs.

632
00:53:14,620 --> 00:53:19,440
They don't need actors and other experiments and some sort of research decides to come along and say,

633
00:53:19,530 --> 00:53:25,270
if you do this, you will get some reward for that long. You as a peace officer, you will get high levels of cooperation, you'll get out of it.

634
00:53:25,470 --> 00:53:31,590
I don't need that to justify their existence, that just their existence is justified by the fact of Facebook behaving in that way.

635
00:53:31,710 --> 00:53:38,520
And of course, you could say again and this is a much wider point, the high levels of customer satisfaction, you better come as customer satisfaction.

636
00:53:38,520 --> 00:53:42,780
Public satisfaction with the police is a positive endpoint in and of itself.

637
00:53:43,110 --> 00:53:47,669
And it's go any further than that. And of course, you don't need experimental designs to trace that.

638
00:53:47,670 --> 00:53:54,240
You can do that with a survey or just by talking to people. And I think that's not just.

639
00:53:55,590 --> 00:53:57,170
Darren, thank you very much.