1 00:00:00,360 --> 00:00:04,630 Okay. Then I was just wondering, would you like to take questions straight after? 2 00:00:04,650 --> 00:00:11,910 Would you want to do it after three pasta dishes and maybe and type two? 3 00:00:11,970 --> 00:00:17,880 Okay. Yeah. Okay. Thanks very much. So thanks for inviting us again, everyone. 4 00:00:17,910 --> 00:00:23,550 It's been a great trip and I talk about experiments in public confidence and ethics that justice this has just been explained. 5 00:00:23,970 --> 00:00:28,860 This is actually a very shortened version of the paper circulated because having written 6 00:00:28,860 --> 00:00:33,240 that paper I realised it would be impossible to present findings from that in 20 minutes. 7 00:00:33,420 --> 00:00:35,940 So this is something that just touched on some aspects of my paper. 8 00:00:36,150 --> 00:00:39,930 If you're interested in this stuff, do have a look at the paper because there's quite a lot more in there. 9 00:00:40,170 --> 00:00:46,260 I'm going to have time to talk about it in the next 20 minutes. In the time I have got to do these things, but it is full of things. 10 00:00:46,530 --> 00:00:48,170 First, give two bits of background. 11 00:00:48,390 --> 00:00:57,630 Contextualise this research in its wider setting, at least a bit in British, Anglo, American, Anglo centric criminology, I think more widely as well. 12 00:00:57,810 --> 00:01:00,209 I'll talk a bit about public confidence in police legitimacy. 13 00:01:00,210 --> 00:01:05,730 Just define some terms to say just define the things I'm talking about and why they're important and why they're interesting. 14 00:01:05,880 --> 00:01:11,850 I'm taking I'm going to introduce a pursuit of justice model and think through the implications for that and the research that I'm presenting. 15 00:01:12,300 --> 00:01:17,130 And then I'm going to look at the causes of trust and legitimacy and say, this is where the experiments aspect of it comes in. 16 00:01:17,370 --> 00:01:21,010 And that's results from two experiments that might Janet are going to talk about. 17 00:01:21,030 --> 00:01:26,790 One is the first one was covered in the paper. If you've had a chance to look at that and I'll finish with some questions. 18 00:01:26,970 --> 00:01:31,020 So this is research and the wider body of research that this this this this particular 19 00:01:31,020 --> 00:01:36,089 presentation addresses is actually interesting in the context of this this conference, 20 00:01:36,090 --> 00:01:40,800 this seminar, because it's mainly about what to say to justice goes right when justice goes right, 21 00:01:40,950 --> 00:01:46,070 or at least what happens what the police and other justice actors can do to make things right. 22 00:01:46,740 --> 00:01:51,240 And actually, it's also an example of when experiments in social research can help. 23 00:01:51,930 --> 00:01:55,140 And hopefully by the end of the presentation, you'll see why they went wrong. 24 00:01:56,010 --> 00:02:00,870 The first bit of background is this the context of this research, in the context of the wider body of research, 25 00:02:00,870 --> 00:02:04,890 which this is about one small part is really in the broadest sense, democratic policing. 26 00:02:05,070 --> 00:02:10,360 And this provides a normative framework. This is the kind of policing that we kind of wants in some sense. 27 00:02:10,360 --> 00:02:14,790 So you could draw on any number of authors to give you an idea about what democratic policing might look like. 28 00:02:15,060 --> 00:02:18,480 Respect, citizens rights, procedural fairness. I'm going to talk about that at some depth. 29 00:02:18,750 --> 00:02:24,030 Minimum use of force, accountability, responsibility, citizen participation, etc., etc. 30 00:02:24,180 --> 00:02:27,630 This is both the kind of policing that we want something everyone wants this. 31 00:02:27,840 --> 00:02:33,030 This is also, I suspect, the kind of policing the research I'm talking about presupposes. 32 00:02:33,270 --> 00:02:39,750 So we've got the police in the United Kingdom are very far from being perfect, but they do at least tick some of these boxes. 33 00:02:39,990 --> 00:02:42,120 And it may be that the things I'm going to be talking about, 34 00:02:42,120 --> 00:02:48,900 this presentation only really make sense when the police services at least take some of these boxes to some extent. 35 00:02:49,260 --> 00:02:53,460 I'll just leave that hanging. That might be something we want to talk about towards the end. 36 00:02:53,730 --> 00:02:59,700 The second bit by background is this, and this is where we're getting into the idea of what can the criminal justice system do to get things right. 37 00:03:00,090 --> 00:03:05,130 So one of the key issues in criminology and we didn't touch too much on it actually in the last couple of days, 38 00:03:05,460 --> 00:03:11,530 but one of the key issues in college is why do people break the law? So what are the factors that determine people's decisions to break the law? 39 00:03:11,550 --> 00:03:17,060 That's the question we usually ask the question in there, in my mind, myself and my colleagues, that none of this is my own work. 40 00:03:17,310 --> 00:03:21,360 I just my own work. This is all very much part of a cognitive project myself. 41 00:03:21,360 --> 00:03:26,430 My colleagues would like to look at this question from a different angle. So we would say, Why do people obey the law? 42 00:03:26,610 --> 00:03:32,070 Why do people make this? That's a positive choice to follow rules and regulations and the law itself. 43 00:03:32,430 --> 00:03:37,280 And of course, there are lots of reasons why people might buy the law, fear of punishment so obvious, 44 00:03:37,320 --> 00:03:41,070 but social motivations, perhaps they might feel guilty to other people if they break the law. 45 00:03:41,340 --> 00:03:44,730 Most people obey most laws most of the time stressing the context, 46 00:03:44,740 --> 00:03:50,100 which I tend to work because I think it's wrong to engage in behaviours that prohibits it in law. 47 00:03:50,400 --> 00:03:57,450 But what the really interesting course of course in this area of research and I think in the in the in the seminar in the last couple of days 48 00:03:57,600 --> 00:04:05,729 is what can justice institutions do to effect law related behaviour and in particular what can justice institutions do to promote law abiding? 49 00:04:05,730 --> 00:04:10,410 This says to not stop people breaking the law, but to encourage people to obey the law. 50 00:04:10,530 --> 00:04:17,280 So just reframing the question, in a sense, the biggest piece of body of work we have around this issue at the moment is 51 00:04:17,280 --> 00:04:20,700 based on the thing called pursuit of justice or the perceived justice model. 52 00:04:20,910 --> 00:04:26,250 It's got a number of names. And this is a model developed by a social psychologist called Tom Tyler in the United States, 53 00:04:26,790 --> 00:04:29,940 starting from the 1980s and going right up to the present day. 54 00:04:30,570 --> 00:04:37,800 And what the procedures this theory of the procedure justice model is, is really a theory, a way of understanding social behaviour in group settings. 55 00:04:38,010 --> 00:04:42,180 So this model is based on the idea that human beings are always acting in group settings. 56 00:04:42,510 --> 00:04:48,630 They're very often, however, acting in group settings where there are unequal power balances within that group system in that context. 57 00:04:48,900 --> 00:04:52,230 So it's obviously particularly relevant in the context of criminal justice 58 00:04:52,410 --> 00:04:56,640 because individuals engaging or acting or talking to whatever they're doing, 59 00:04:56,880 --> 00:04:59,720 just as actors are doing, do so across quite a steep. 60 00:05:00,080 --> 00:05:07,040 Power imbalance, the police officer or the judge or whoever it is, has a lot of power over the individual citizen in almost every case. 61 00:05:07,400 --> 00:05:14,299 And what the pursuit of justice photo does is provide a way of understanding what types activity, police officers, for example. 62 00:05:14,300 --> 00:05:19,730 And I'm going to limit today's conversation to police officers. It could apply to any other criminal justice act as well, 63 00:05:19,940 --> 00:05:25,850 provides a way of understanding what types of police behaviour might promote or inhibit compliance and cooperation. 64 00:05:26,000 --> 00:05:31,100 So what might encourage people to obey the law rather than prevent people from watching and pursuit of justice? 65 00:05:31,310 --> 00:05:37,820 I think like all good ideas, it's got at its heart a really simple, basic observation, and that's in the context of others, 66 00:05:37,820 --> 00:05:41,600 such as the ones I've been talking about, where we have unequal power relations in groups settings. 67 00:05:41,810 --> 00:05:46,610 What people value in their interactions with power holders in those settings is just indecent treatment, 68 00:05:46,820 --> 00:05:49,820 transparent, fair decision making, sense of respect, 69 00:05:50,000 --> 00:05:57,020 a sense of equality within that relationship, at least within that moment, when people feel treated in this way by just as actors, by police officers. 70 00:05:57,230 --> 00:06:03,050 And this promotes a sense of exactly justice that this person with whom I was acting is acting in a fair way. 71 00:06:03,290 --> 00:06:07,249 It promotes a sense of trust in this actor and their institutional organisation. 72 00:06:07,250 --> 00:06:11,630 They represent just pursuit of justice. Trust in turn, enhance legitimacy. 73 00:06:11,930 --> 00:06:16,670 Enhance legitimacy, leads to decision acceptance, cooperation and compliance with the law. 74 00:06:17,070 --> 00:06:19,220 This is the core claim of the pursuit of justice. 75 00:06:19,580 --> 00:06:25,220 In second core claim running alongside this is these processes are more important than instrumental concerns. 76 00:06:25,670 --> 00:06:28,130 So when people are thinking about the legitimacy of the police, 77 00:06:28,280 --> 00:06:33,110 their thinking much more about the procedural fairness of police, whether they trust the police in this sense, 78 00:06:33,260 --> 00:06:39,650 than they are about whether the police are effective in dealing with crime and in combating in capturing offenders and prosecuting offenders, 79 00:06:39,800 --> 00:06:43,370 etc., etc. It's not to say they don't care about those things because they do. 80 00:06:43,550 --> 00:06:47,030 They just care more about the fairness of interactions they have with officers 81 00:06:47,150 --> 00:06:50,690 or interactions they imagined they might have with officers in the future. 82 00:06:51,350 --> 00:06:54,530 What is the pursuit of justice worth? Why does it have this effect? 83 00:06:54,830 --> 00:06:58,219 Well, I think remember, it's all based on group settings. 84 00:06:58,220 --> 00:07:01,520 What procedural fairness does is when you're interacting with a police officer, 85 00:07:01,760 --> 00:07:06,440 that police officer represents to you an important social group might characterise it as a nation state. 86 00:07:06,470 --> 00:07:12,170 Obviously there's a lot of material in sociology. Police have says that's exactly what the police represent to most people. 87 00:07:12,350 --> 00:07:16,310 When that officer treats you with fairness, with dignity and respect, you feel included. 88 00:07:16,460 --> 00:07:20,240 You feel you belong, you feel you have status within the group the peace officer represents. 89 00:07:20,450 --> 00:07:24,860 When you feel unfairly treated, you feel excluded, you feel pushed out and denigrated. 90 00:07:25,760 --> 00:07:31,880 And when you feel inclusion, the group you're motivated to as authorities, you're motivated to follow its rules. 91 00:07:32,030 --> 00:07:35,809 When you feel excluded from a group and you feel pushed out, they're not motivated. 92 00:07:35,810 --> 00:07:39,530 They're just sponsors of forces. You're not motivated to abide by those rules. 93 00:07:39,770 --> 00:07:45,080 Basically, if a group is excluding you, why should you off going along with what the rules that govern the behaviour of group? 94 00:07:45,560 --> 00:07:48,650 And I think there's an important other aspect of Syria. 95 00:07:48,650 --> 00:07:55,010 So another component and a part of its effects lies in the idea that police should be exemplars of good conduct. 96 00:07:55,370 --> 00:07:59,839 So when a police officer acts in a way that you think they should not back in of was treating with dignity, 97 00:07:59,840 --> 00:08:03,350 respect or treating other people around you with dignity and respect, you think, 98 00:08:03,350 --> 00:08:07,430 well, he's fulfilling his side of the bargain or she I should feel fulfilled myself. 99 00:08:08,210 --> 00:08:12,230 When a police officer doesn't act in those ways, you're going to think, well, they're not doing the right thing. 100 00:08:12,230 --> 00:08:18,110 Why should I bother doing the right thing? So I think pursuit of justice has an effective component and it has an evaluative component. 101 00:08:18,290 --> 00:08:24,140 And both these things feed into legitimacy judgements. People make feed into cooperation, compliance and so on. 102 00:08:24,350 --> 00:08:28,999 And if this model is right, of course cooperation and compliance of the law can be secured, 103 00:08:29,000 --> 00:08:33,760 most importantly by justice agents, by police officers, not by demonstrating instruments like most of us. 104 00:08:33,760 --> 00:08:38,960 I don't reiterate that is important, but primarily by treating people with dignity and respect, 105 00:08:39,470 --> 00:08:43,590 by making them feel included, making them feel like long in the group the police represent. 106 00:08:44,750 --> 00:08:46,790 I talked a lot about legitimacy. 107 00:08:46,790 --> 00:08:52,550 Just just a brief thought about what I mean by legitimacy and in this context is quite important only from a legal context. 108 00:08:52,760 --> 00:08:55,760 I'm dealing with an empirical notion of legitimacy here. 109 00:08:55,940 --> 00:09:00,470 So I'm not thinking about things that actually make the police legitimate in some objective sense. 110 00:09:00,650 --> 00:09:04,639 I'm talking about the way that people react to social institutions and the extent to which 111 00:09:04,640 --> 00:09:09,230 they legitimise those social institutions based on or not based on their own opinions, 112 00:09:09,230 --> 00:09:13,790 experiences, orientations and so forth. If you see legitimacy in these terms, 113 00:09:13,970 --> 00:09:17,540 you can find a number of ways you can say it's property and authority or 114 00:09:17,540 --> 00:09:22,130 institution that leads people to feel is entitled to be deferred to and obeyed. 115 00:09:22,280 --> 00:09:28,490 That's also important in criminal justice context that nevertheless, the argument here would be when someone grants the police legitimacy, 116 00:09:28,640 --> 00:09:33,470 they feel I have a duty to abide by the instructions of police officers and then grant police such as mercy. 117 00:09:33,590 --> 00:09:35,180 They don't feel any sort of sense of duty. 118 00:09:35,330 --> 00:09:40,820 And when they're at such a sense of duty as absence, police have to rely on force to get what they want out of people. 119 00:09:41,300 --> 00:09:43,190 You can get slightly more expansive. 120 00:09:43,190 --> 00:09:49,370 You can say legitimacy is a right to enforce commands which cannot be countermanded as have a monopoly of such legitimate force to. 121 00:09:49,370 --> 00:09:54,860 Hey, we're going back to favour the idea that the police are the monopolist of legitimate force within society, 122 00:09:55,430 --> 00:09:59,450 and we in this work tend to use a slightly more and even more, if you like, expansive. 123 00:09:59,800 --> 00:10:03,040 Notion of legitimacy is based on what physical scientists call. They be firm. 124 00:10:03,190 --> 00:10:10,600 And he's, got to my mind, still the best definition of best understanding the legitimacy of the way we're thinking about justice in this context. 125 00:10:10,960 --> 00:10:14,740 And he thinks legitimacy is always going to sit on the basis of common shared values. 126 00:10:14,980 --> 00:10:18,070 This is actually a quietly radical idea when you think about it. 127 00:10:18,190 --> 00:10:23,110 This means that the legitimacy of important state institutions such as APC is always, in a sense, up for grabs. 128 00:10:23,500 --> 00:10:27,730 It's always in if, say, of change of state flux to have more or less a different context. 129 00:10:27,880 --> 00:10:33,850 People make judgements about police. It goes into de-legitimize case based on their own experiences they're dismissing. 130 00:10:33,850 --> 00:10:38,620 The place is not a given. It's always in flux and it goes beyond that history. 131 00:10:38,620 --> 00:10:44,169 Dimensions must be fulfilled for power to be considered legitimate conformity to a set of rules to justify that a 132 00:10:44,170 --> 00:10:49,420 city's rules in terms of shared beliefs and the express consent of those governed or often affected by the past, 133 00:10:49,670 --> 00:10:52,630 just about conforming to the rules that govern the behaviour of the institution. 134 00:10:52,900 --> 00:10:57,820 It's about judgement of whether those are the right rules and it's about the expressed consent of the people. 135 00:10:57,820 --> 00:11:06,730 So it's not naive to believe that police are just basically just they have to acting as if they had to express and it just mixing up ideas as it were. 136 00:11:07,900 --> 00:11:11,080 So this would be one way of kind of conceptualising some of what I've just said. 137 00:11:11,260 --> 00:11:16,660 So on the left are the same assets. On the left hand side, we've got factors that feed into trust in police. 138 00:11:17,020 --> 00:11:24,280 So I've spent a lot of time talking about personal contact and imagining personal contact experiences of policing to personal contact with officers, 139 00:11:24,280 --> 00:11:30,430 vicarious contacts, the kind of conversations, ideas or stories about the police that circulate social groups. 140 00:11:30,670 --> 00:11:34,600 You can put the major and they can be the have for people's cultural repertoire. 141 00:11:34,600 --> 00:11:36,970 They just imagining that brings the police. 142 00:11:37,180 --> 00:11:43,749 You can have a course experience of crime victimisation and social behaviour have social psychological factors, authoritarianism. 143 00:11:43,750 --> 00:11:46,790 Some people are just motivated to grant legitimacy to authority figures. 144 00:11:46,790 --> 00:11:50,750 They just like it in some sense system justification power to their decision. 145 00:11:50,920 --> 00:11:52,299 This is basically just bringing home the idea. 146 00:11:52,300 --> 00:11:59,020 There are lots of factors going on behind driving trust in police driving legitimacy that flows from that trust. 147 00:11:59,510 --> 00:12:05,820 I'm not that that's what I'm claiming happens. I'm going to concentrate today on just one very specific aspects of this wider motive, 148 00:12:05,860 --> 00:12:09,700 which is personal contact between individuals and police officers. 149 00:12:09,850 --> 00:12:15,459 And look, we look to the extent to which we can say the personal contact in some sense causes trust and in some 150 00:12:15,460 --> 00:12:20,800 sense causes legitimacy as a causal link with these ideas and motivations that people might have. 151 00:12:21,700 --> 00:12:28,690 And as I've kind of intimated at the beginning, pursuit of justice is fair, is primarily it's been developed in Anglophone contexts, 152 00:12:29,170 --> 00:12:34,329 the US, UK, Australia, etc. It's finding increasing purchase in the over developed world. 153 00:12:34,330 --> 00:12:41,260 If you want to use that phrase, it's much less clear whether it works in non Anglophone contexts or in the Global South. 154 00:12:41,500 --> 00:12:46,059 I'm going to leave that question hanging with returns by the end of the question I do want to 155 00:12:46,060 --> 00:12:51,040 address is that most evidence on these issues and basically almost everything I've said so far, 156 00:12:51,730 --> 00:12:55,690 the evidence for that primarily comes from cross-sectional sample surveys, 157 00:12:55,900 --> 00:13:00,850 opinion poll research, basically, which is my way of saying that the evidence we have, 158 00:13:00,850 --> 00:13:04,059 the kind of relationships I've been talking about of good is good, 159 00:13:04,060 --> 00:13:08,740 but it's not that good because it's only based on based on these cross-sectional sample size. 160 00:13:08,890 --> 00:13:13,270 I mean, that's a slight caricature. The research is slightly more rich than that. 161 00:13:13,350 --> 00:13:19,570 But what I want to talk about today is field experiments or the conduct of a field experiment. 162 00:13:19,750 --> 00:13:24,550 And that allows us to test whether the experience of procedural fairness at the hands of police officers 163 00:13:24,730 --> 00:13:30,070 really does cause really is causally linked to people's trust judgements and their legitimacy judgements. 164 00:13:30,160 --> 00:13:36,370 And then we would think cognitive cooperation complies etc. etc. because 165 00:13:36,430 --> 00:13:41,800 cross-sectional data survey data can uncover association that can uncover causation. 166 00:13:42,120 --> 00:13:48,999 I think coming to labour that point experiments of the type that we're talking about now do allow assignment of causal 167 00:13:49,000 --> 00:13:54,430 effects and I won't go too much into the kind of sexualising experiments of it's fairly obvious what they are. 168 00:13:54,640 --> 00:13:59,560 But one way of thinking about what an experiment is in this context is, is to look at the work of studies. 169 00:13:59,650 --> 00:14:03,040 So they say or all experiments or social experiments. 170 00:14:03,370 --> 00:14:08,980 Experiments actually in general have four elements that they share in common a variation in treatment. 171 00:14:09,260 --> 00:14:13,879 A You do something to one person or one group of people who have experiments. 172 00:14:13,880 --> 00:14:17,380 The subjects are you don't do it to another person or another group of people. 173 00:14:17,740 --> 00:14:24,490 They have post-treatment measures of outcomes. You measure stuff. In the experiment, you have at least one unit in which the observation is made. 174 00:14:24,730 --> 00:14:32,830 I think in our case, people and you have a mechanism for assessing what the outcome would have been had you had it not been for the treatment, 175 00:14:32,950 --> 00:14:39,669 which is just a fancy way of saying you have a treatment group, have an experimental group that get the thing that you think has the causal effect on 176 00:14:39,670 --> 00:14:43,720 the outcomes you're interested in and you have a control group don't get that thing. 177 00:14:43,960 --> 00:14:48,430 You comparing between the two and that comparison between the two allows you to exacerbate the 178 00:14:48,430 --> 00:14:54,640 causal effects somewhat insists that true experiments require randomisation of the treatment. 179 00:14:54,940 --> 00:14:59,540 I'm actually experiment I'm going to explain to you now does have randomisation of treatment not all it's. 180 00:14:59,560 --> 00:15:03,760 Experiments do need randomisation, actually. And the second experiment is in the paper. 181 00:15:03,970 --> 00:15:08,020 It's a non-randomized experiment, what's known as quasars. But that's just inside. 182 00:15:09,000 --> 00:15:15,360 So I tried to claim I think the contact, personal contact with police officers is probably one of the strongest influences 183 00:15:15,570 --> 00:15:18,900 on people's trust judgements on the legitimacy that they brought to the police. 184 00:15:19,860 --> 00:15:26,880 The best evidence we have for causal association here so far comes from something called the Queensland Community Engagement car or queue set. 185 00:15:28,440 --> 00:15:32,729 This does provide experimental evidence that the quality of interaction between police 186 00:15:32,730 --> 00:15:37,800 officers and individuals has a causal direct effect on perceptions of police fairness, 187 00:15:38,040 --> 00:15:41,130 satisfaction of encounters, trust the police legitimacy. 188 00:15:41,430 --> 00:15:47,620 But I do think that what we conducted in Queensland, as the name suggests in Australia in 2010, 189 00:15:47,640 --> 00:15:55,470 I think they took wrote random roadside breath tests of which they did a huge amount in Australia, millions of these things each year. 190 00:15:56,220 --> 00:16:01,870 The baseline encounter in these road tests between an officer and a citizen is very, very quick. 191 00:16:01,890 --> 00:16:06,600 Basically they just pull the car in a match up to the window. They say blowing this, you are negative. 192 00:16:06,600 --> 00:16:11,700 Off you go. Encounters lost less than 30 seconds on that basis for the Q set child all do. 193 00:16:11,790 --> 00:16:19,769 They took some of the core principles of procedural justice, dignity, respect, clear communication, transparency, etc., 194 00:16:19,770 --> 00:16:28,050 etc. They operationalise those into a checklist that the officers had to deliver in this encounter increased in council length to 90 seconds, 195 00:16:28,260 --> 00:16:34,980 and as a result, the people who received that treatment, received that inspection intervention, had high levels of trust in police fairness, 196 00:16:35,280 --> 00:16:40,380 higher levels of satisfaction with the encounter, and they granted the police more legitimacy on that basis. 197 00:16:41,010 --> 00:16:44,820 The experiment I want to talk about was designed as exactly as a replication of this. 198 00:16:44,970 --> 00:16:49,800 So one of the key strands, I think the experimental method is you can replicate it across different contexts 199 00:16:49,920 --> 00:16:53,129 and actually it's in the process of replication that you really be able to get. 200 00:16:53,130 --> 00:17:00,930 Firm explores the causal effects that you're interested in. And this was conducted in Scotland and the Christmas, Christmas just past. 201 00:17:01,620 --> 00:17:05,130 And we tried to do it myself and my colleague Sarah McLean from the University of Edinburgh. 202 00:17:05,250 --> 00:17:12,030 We tried to do it in as close the way or way that matched what they'd done in Australia as closely as possible. 203 00:17:13,050 --> 00:17:17,910 So this was during the first National Festive Road Safety campaign and that's what it says on the tin really. 204 00:17:17,910 --> 00:17:23,470 It's all about drink driving safety at Christmas period and to try to encourage drivers to not drink, 205 00:17:23,490 --> 00:17:28,709 drive safely, etc., etc., they don't do random stops, they claim. 206 00:17:28,710 --> 00:17:33,540 I think they've got the duty random stops in effect. But this is slightly different context. 207 00:17:33,540 --> 00:17:38,879 Drivers are stopped by the police with the aim of getting drunk driving and primarily 208 00:17:38,880 --> 00:17:43,860 actually improving vehicle and driver safety so that somehow they've got broken tail lights. 209 00:17:44,400 --> 00:17:48,570 They may or may not give them a ticket for the broken taillight. They'll have a chat with them. 210 00:17:48,750 --> 00:17:53,550 They say You need to give her a light fix. If they think that's a good reason, they'll give them a breath test. 211 00:17:53,700 --> 00:17:59,400 You get the idea. It's kind of it's kind of an instrumental intervention to strike fear into drivers, 212 00:17:59,550 --> 00:18:04,590 but it's also kind of a process based intervention to try and develop links between the police and community, 213 00:18:04,720 --> 00:18:09,330 try to demonstrate to the community that police care about their safety and the safety of other people on the roads, 214 00:18:09,330 --> 00:18:14,220 etc., as around 20,000 stops in Scotland over this period. 215 00:18:14,850 --> 00:18:18,570 And we did we utilised a matched pairs three post design. 216 00:18:18,870 --> 00:18:22,830 So basically road traffic policing in Scotland is basically almost two units. 217 00:18:23,070 --> 00:18:27,090 There are 20 of these units with group them. It's pairs based on similarity. 218 00:18:27,330 --> 00:18:31,830 So for example Glasgow and Edinburgh compared because they're the two largest cities as similar. 219 00:18:31,830 --> 00:18:35,730 All the other areas were pattern by similarity. So we had ten pairs. 220 00:18:36,060 --> 00:18:41,490 We then literally flipped a coin to decide which one of the pair got the intervention, which one of the pair didn't. 221 00:18:41,910 --> 00:18:48,060 Then we had a pre survey, so all the officers in made stops in the pre survey period handed out fishnet people. 222 00:18:48,570 --> 00:18:52,320 We then had a change over period where we provided them with some briefing period materials. 223 00:18:52,620 --> 00:18:58,980 Then the post period they did exactly the same. They implemented the intervention in the test areas and they handed out a questionnaire. 224 00:18:59,400 --> 00:19:02,070 So pre and post measures of opinions. 225 00:19:02,220 --> 00:19:09,600 We have intervention in the test areas and we have the control areas and again comparison of what changes between pre and post 226 00:19:09,600 --> 00:19:15,780 in tested control areas will give us an idea of whether the intervention we designed really did increase trust and legitimacy. 227 00:19:18,150 --> 00:19:19,020 We suffered. 228 00:19:19,040 --> 00:19:26,130 We were experiencing difficulties in design, office and business as usual in Scotland was was different to business as usual in Australia. 229 00:19:26,340 --> 00:19:33,360 Remember in Australia the encounters were very short, 30 seconds and in Scotland they were much longer, much more discursive. 230 00:19:33,360 --> 00:19:40,200 The officers and basically the officers were already implementing a lot of justice in the encounters they were doing, as it were. 231 00:19:40,290 --> 00:19:42,510 They were good at explaining to people why had been stopped. 232 00:19:42,690 --> 00:19:46,410 They were good at managing those clients as a way to demonstrate some respect and dignity. 233 00:19:46,590 --> 00:19:51,300 So we had a problem here of implementing an intervention that was significantly different to business as usual. 234 00:19:51,480 --> 00:19:56,340 But we tried. So what we just did is again designed a checklist which had a number of key elements 235 00:19:56,340 --> 00:20:01,020 the officers were asked to try and get across in the process every very who said, 236 00:20:01,020 --> 00:20:04,469 please do all of these things at the moment. You probably do some of them. 237 00:20:04,470 --> 00:20:08,370 Can you do all of them at the moment? And we also included a leaflet to try to. 238 00:20:08,600 --> 00:20:13,759 Blind to people why this festive campaign was running, what the police were going to get out of it, 239 00:20:13,760 --> 00:20:18,050 what the public should get out of it in terms of increased road safety, etcetera, etcetera. 240 00:20:18,200 --> 00:20:22,350 And the key message was were around respect, equality, trustworthy motives. 241 00:20:22,370 --> 00:20:29,180 I explained why people why they'd been stopped. Dignity, neutrality, all the core aspects of pursuit of justice theory. 242 00:20:29,600 --> 00:20:31,580 And that's just a picture of, you know, you can't see that. 243 00:20:31,700 --> 00:20:37,129 That's just a picture of the leaflet that we designed with Police Scotland to hand out to everyone in the test group. 244 00:20:37,130 --> 00:20:43,550 Got one of these leaflets, leaflets, everyone the control group did not for basic hypotheses. 245 00:20:43,550 --> 00:20:49,070 We wanted to test. We wanted to see whether the intervention or this procedure of justice intervention that we tried to sign, 246 00:20:49,760 --> 00:20:56,540 increased perceptions of pursuit of justice, doing the stop itself, whether the increased satisfaction with the stop did it. 247 00:20:56,540 --> 00:20:59,390 Were you satisfied with the way the officer treated you on this occasion, 248 00:20:59,690 --> 00:21:04,350 whether increased trust and confidence in the police and whether increased police legitimacy? 249 00:21:04,380 --> 00:21:10,520 So for our purposes. Just a quick slide on the kind of use of randomisation that the vaguely evangelical 250 00:21:10,520 --> 00:21:13,669 bits of this I'm not particular I'm not particularly evangelical about experiments. 251 00:21:13,670 --> 00:21:17,840 I think in some respects are deeply problematic, but they are very useful if you're into social research. 252 00:21:18,080 --> 00:21:24,379 So we got a very low response rate in this place. So as you usually get with Postal Service, if you've got less than 8% of the service, 253 00:21:24,380 --> 00:21:28,840 but it's still a sample size of about 8800 or something, which is plenty. 254 00:21:29,280 --> 00:21:34,340 But because of the randomisation that doesn't matter because the treatment was randomised on the flip a coin, 255 00:21:34,550 --> 00:21:38,780 the actual response rate is immaterial and since it's a non-representative sample, 256 00:21:38,780 --> 00:21:44,180 we can't claim that this sample is even representative of the people who are out in Scotland's roads over this period. 257 00:21:44,600 --> 00:21:52,190 That's not the point. It doesn't have to be a randomised representative sample again because the randomisation process this is 258 00:21:52,190 --> 00:21:57,230 this is why one of the reasons why experiments are good as well as being able to cite causal effects. 259 00:21:57,470 --> 00:22:05,330 But what we found is overall across treatment and control groups, driver perception, the police encounter was very, very positive. 260 00:22:05,330 --> 00:22:10,730 So this basically confirms what we thought. Originally police in Scotland were already very good at doing this, 261 00:22:11,120 --> 00:22:15,859 so at least meant for one thing and this is what we kind of suspected from beginning our intervention was 262 00:22:15,860 --> 00:22:19,960 rather weak because it wasn't very different from what they were doing at the moment compared to trust, 263 00:22:19,970 --> 00:22:23,580 Australia was in fact strong. It was quite different what in English. 264 00:22:23,990 --> 00:22:29,600 But what we found was perceptions of procedural justice increased in the control group that didn't 265 00:22:29,600 --> 00:22:34,250 get the intervention and they failed in the experimental group which did get intervention. 266 00:22:34,520 --> 00:22:38,470 So it's not that we didn't have any effect. Of course we made things worse by the intervention. 267 00:22:38,660 --> 00:22:44,899 This is well talking about research going wrong and no effects on trust satisfaction with the encounter. 268 00:22:44,900 --> 00:22:50,810 Did you like the way the police officer treated you overall? Went up in the control group and went down in the experimental group. 269 00:22:51,050 --> 00:22:55,510 Again, we made things worse rather than better, and the legitimacy was not effective. 270 00:22:55,720 --> 00:23:01,340 It's not that I'm sure if we go back to our hypothesis, that's just legitimate. 271 00:23:01,340 --> 00:23:10,110 So I'm okay with that. I'll go back to, I hope, the intervention damaged perceptions of trust, genuine cancer engagement, decreased satisfaction. 272 00:23:10,140 --> 00:23:17,210 The stock had no effects on trust and confidence overall in the police had no effects on the legitimacy of police in some overall sense. 273 00:23:17,780 --> 00:23:23,750 So why didn't it work? Well, we're still engaged in some and some research around why it didn't work, why we made things worse. 274 00:23:23,930 --> 00:23:29,660 But I think we can make some educated guesses. One thing we think we did was bureaucratised the encounter. 275 00:23:29,990 --> 00:23:36,110 So we made the officers so concerned with filling in well, not filling in, but meeting all the criteria that been given in that checklist. 276 00:23:36,290 --> 00:23:42,440 They forgot to manage the encounter as human beings. We slightly turned them into robots, if you like. 277 00:23:42,890 --> 00:23:46,310 I say these effects are small, but they are statistically significant. 278 00:23:46,610 --> 00:23:52,370 We also suspect that we made the process longer. Again, they were so concerned with doing all the things we asked them to do. 279 00:23:52,580 --> 00:23:55,549 The stops became longer. They became more intrusive. 280 00:23:55,550 --> 00:24:01,910 On that basis, they may even have been perceived experience as procedurally unfair by the people concerned because they were longer. 281 00:24:02,150 --> 00:24:06,980 So if you if you drag things out, you're kind of suggesting to people you don't really respect them of their time, 282 00:24:07,280 --> 00:24:10,410 for example, and you're not worried that they might have somewhere to go. 283 00:24:10,440 --> 00:24:15,500 Want to get this over to move as quickly as possible? It's also possible the effects of the leaflets were negative. 284 00:24:16,310 --> 00:24:18,290 It may be something in the design of the leaflet. 285 00:24:18,840 --> 00:24:23,840 It may be something about reminding people that some have been accused of being bad drivers and have been stopped. 286 00:24:23,840 --> 00:24:27,559 That might have the effect. We don't know about that. To say we do some follow up study. 287 00:24:27,560 --> 00:24:32,720 The follow up site is trauma cop will have I would say having effects at all is great. 288 00:24:33,170 --> 00:24:38,960 So this is actually I was sceptical right through this process that we find any results from this experiment whatsoever. 289 00:24:39,020 --> 00:24:44,310 I really didn't think we bad to have. In fact we did have a fact just because it wasn't the direction that we kind of wanted. 290 00:24:44,420 --> 00:24:49,280 It doesn't negate the importance that we found in effect and actually working at what happened provide a lot of 291 00:24:49,280 --> 00:24:54,380 information for in this case the Scottish Police and Scottish Government on what they might need to do in the future. 292 00:24:55,220 --> 00:24:58,550 Just lastly, things to to to clear up on that. 293 00:24:59,900 --> 00:25:03,229 Some things can go right or wrong as well as right. 294 00:25:03,230 --> 00:25:08,360 When you do this kind of research, you really are kind of, if you like, putting your theories, putting ideas right on the line. 295 00:25:08,360 --> 00:25:12,350 You. To be proved wrong, that kind of kind of absolute sense. 296 00:25:13,400 --> 00:25:17,870 How do you deal with these perverse outcomes theoretically in policy terms? 297 00:25:18,200 --> 00:25:23,060 So when I come back to this point in a minute, but one policy implication of this, if you interpret it in the wrong way, 298 00:25:23,180 --> 00:25:27,020 would be the police don't really need to care about justice when they're already doing all this stuff. 299 00:25:27,020 --> 00:25:31,940 Fine. They've got no issues with this whatsoever. Well, I can tell you in the context of policing and asking them that isn't true. 300 00:25:32,150 --> 00:25:37,340 They do have quite significant concerns and problems around implementation of procedures and policing. 301 00:25:37,520 --> 00:25:40,640 But you could use the results of this experiment that we don't need to worry about this stuff. 302 00:25:40,850 --> 00:25:44,530 How do you deal with that? It's almost a presentation issue, a bit deeper. 303 00:25:45,200 --> 00:25:51,410 Experiments don't really work in isolation. So we might conclude on the basis of the case that this is fantastic. 304 00:25:51,530 --> 00:25:57,080 All we need to do is increase the length of encounters, get this office to tell people what they're doing and trust the legitimacy will increase. 305 00:25:57,230 --> 00:25:58,670 All right. So that's actually no. 306 00:25:58,850 --> 00:26:04,910 You need to be very careful about the kind of implementation, the kind of interventions you design and how you implement them. 307 00:26:05,930 --> 00:26:09,740 Beyond that, I would say replication against recipient replication is a really powerful device. 308 00:26:09,890 --> 00:26:13,820 It's never good enough to do this kind of research once or twice. You got to do it lots of times. 309 00:26:13,910 --> 00:26:20,510 It's a bit of a body of knowledge, if you like. And of course, leading off from that kind of problems is experiments, not panacea. 310 00:26:20,630 --> 00:26:24,140 So it's not kind of positive about this idea, I think is a really good way of doing research. 311 00:26:24,290 --> 00:26:28,639 It's not going to provide you with all the answers to all your problems or real questions and can create as 312 00:26:28,640 --> 00:26:35,030 many problems in as many questions as it so complex interventions that the part we often interested in, 313 00:26:35,210 --> 00:26:39,800 in criminal justice, are very different, difficult to implement in the experimental context. 314 00:26:40,010 --> 00:26:44,899 So even when the Scot said, it seemed very simple and I invented it as it seems very simple. 315 00:26:44,900 --> 00:26:52,610 I think actually it was a quite a complex intervention and as much as we had the test, the checklist we've given to people and we have the leaflet. 316 00:26:52,790 --> 00:26:54,560 So there's actually two interventions wrapped up in one. 317 00:26:54,680 --> 00:27:01,610 If we'd have had a positive effect with a negative effect, we don't know whether it was the checklist we'd given to the officers or have that effect. 318 00:27:01,850 --> 00:27:07,010 We don't know what would have happened if we take one away would be it's got a lot of effect we observe, but we got another effect. 319 00:27:07,310 --> 00:27:10,340 Gratification in the future can help solve these types of problems. 320 00:27:10,490 --> 00:27:17,690 And that that's two interventions. And you think about justice intervention or an intervention, of course, 321 00:27:17,810 --> 00:27:23,960 they're going off to have very many different aspects can be very difficult to design experiments that can address those types of issues. 322 00:27:24,680 --> 00:27:30,020 There's always a problem, and this is best summarised by Nancy Cartwright, and I can't remember surname Heidi, 323 00:27:30,020 --> 00:27:34,820 that the problems of experiments and previous social research is that it works there. 324 00:27:34,850 --> 00:27:38,390 But will it work? So we found it works in Queensland. 325 00:27:38,690 --> 00:27:42,670 We found it didn't work in Scotland and it works there, but it's good. 326 00:27:42,800 --> 00:27:49,940 Will it work here? It's also a problem of implementation. This experiment, this A.C.T. results suggest that this implementation is good. 327 00:27:50,060 --> 00:27:54,990 How do we get, from the context of the experiment itself to the wider embedding of that post, 328 00:27:55,010 --> 00:27:58,400 a change in the concepts of actually working within which you're working? 329 00:27:59,260 --> 00:28:04,370 And a final thought, I mean, I think one and this is tapping into some of the bigger issues around this approach, 330 00:28:04,580 --> 00:28:08,600 certainly within British criminology and American College at the moment, 331 00:28:08,810 --> 00:28:17,410 I think there's an increasing preponderance on the experiments, measures, methods in some sections of criminology and broadly that's to be welcomed. 332 00:28:17,420 --> 00:28:22,100 It is a methodological advance over the methods and approaches we've been using before. 333 00:28:22,880 --> 00:28:27,020 And it's great because you can assign causal facts and causes of what we used to interest you. 334 00:28:27,230 --> 00:28:31,700 It comes with a number of attendant risks, which I spend a lot of time talking about in the paper. 335 00:28:31,880 --> 00:28:35,720 But just to summarise the one which I think is probably the most important for me, 336 00:28:35,900 --> 00:28:39,680 I think this approach risks instruments for those in criminal justice research. 337 00:28:42,180 --> 00:28:49,320 Basically you can claim that procedurally that you should claim procedurally fair procedural justice is a good in itself. 338 00:28:49,650 --> 00:28:54,240 It doesn't matter if it purchases trust, if you like. It doesn't matter if it generates legitimacy. 339 00:28:54,450 --> 00:28:57,690 Those are that's just the way that police officers should treat people. 340 00:28:57,960 --> 00:29:03,750 And it just needs rely on the idea that somehow works to generate trust and legitimacy to save police this issue. 341 00:29:03,870 --> 00:29:09,240 This is what you should be doing. The danger of experimental approaches are we can say, well, actually it doesn't have much effects. 342 00:29:09,440 --> 00:29:12,690 If this is too complicated a social environment for us to expect, 343 00:29:12,690 --> 00:29:17,520 any particular intervention or pursuit of justice will generate an effect in terms of trust and legitimacy. 344 00:29:17,610 --> 00:29:20,910 We could then, on that basis just say, well, we don't need to worry about perceived justice, 345 00:29:21,000 --> 00:29:24,420 or we can't worry about perceived justice because it seems it doesn't have an effect. 346 00:29:24,660 --> 00:29:30,720 That's the danger. And another way of rephrasing that question and I'll finish here, I'm high customer satisfaction. 347 00:29:31,200 --> 00:29:35,610 It's an inverted commas round. It is again a good in and of itself. 348 00:29:36,090 --> 00:29:37,260 We don't need an experiment. 349 00:29:37,260 --> 00:29:43,350 We didn't need an experiment to survey drives in Scotland and find that they were generally very happy with the examples they encountered, 350 00:29:44,040 --> 00:29:47,350 that we could have an unbiased survey, we could have just done that by talking to. 351 00:29:47,520 --> 00:29:53,790 So in that there was a way of saying that there are lots of research questions that we might be interested in that don't need. 352 00:29:53,790 --> 00:29:58,740 The kind of approaches I've been talking about today and the findings of those research projects can be just as important, 353 00:29:58,890 --> 00:30:02,100 if not more important than the kind of things I've been talking about now. 354 00:30:02,370 --> 00:30:05,940 And I will finish and I think. Yes. Thank you very much. 355 00:30:06,240 --> 00:30:11,550 Thank you. Thank you.