Version classiqueVersion mobile

Tyneside Neighbourhoods

Daniel Nettle

6. Being there

Texte intégral

The places you play and where you stay,
Looks like one great big alleyway.


1As I mentioned in chapter 1, the Tyneside Neighbourhoods Project was not just about how behaviour varies from neighbourhood to neighbourhood, but also about the psychological mechanisms underlying this variation. I hope you will grant me, from all the data presented so far, that social behaviour differs profoundly from Neighbourhood A to B. I hope you will also grant me that a plausible explanation for this is that living in Neighbourhood A affects people in specific ways that are different from the ways living in Neighbourhood B affects the people who live there. That is, I assume that the behavioural differences between A and B are to a considerable extent consequences of living in those places, rather than being due some other exogenous factor, such as differential migration of people with certain personalities to certain neighbourhoods.

2This raises a fundamental question: how does your psychology come to be affected by the community you live in? The classic anthropological answer invokes acculturation, as if no further elaboration was required: in the words of Herder quoted by Morin (2015), people just absorb their culture like ‘a wet sponge that has long been soaking on a wet floor’. The more you examine this metaphor, the more you realise that invoking acculturation without further unpacking does not explain anything. Things outside the mind (acts, artefacts, etc.) cannot literally move inside the mind as the water can move into the sponge. The most that can happen is that the mind can construct certain kinds of internal representations (beliefs, memories, attitudes) in a way that is driven by cues in its environment. These cues might — or might not — be there in the local environment because of the behaviour of other members of the community, behaviour that might in turn be driven by the beliefs, memories, and attitudes that they hold. Thus, when we say that you can acquire attitude x from others in your community through acculturation, what we mean is that: (1) the attitude x held by others causes them to emit certain classes of environmental cues, (2) these cues can be detected, and (3) these cues increase the probability of your mind constructing x rather than some alternative attitude. This immediately raises more fascinating and more tractable questions. What are the cues? How do we detect them? What timescale do they operate on?

3Scholarly discussions of acculturation also tend to overstate the importance of simple imitation, as if cultural traditions were reducible to people copying the acts of others in their surrounding communities. The problem with this is that it does not account for the open-endedness of behaviour; people can come up with appropriate and socially-patterned decisions for situations they have never faced before, and never seen anyone else face either. For example, when participants were asked to play Kari Britt Schroeder’s Theft Game (chapter 4), they had never played that game or even been in a closely similar situation before. Thus, there is no way they could have been following a rule of the kind ‘behave in this game as I have seen others from my community behave in this game on previous occasions’. Yet they came up with coherent patterns of responses that were systematically different across Neighbourhoods A and B.

4Instead, we must assume that they first parsed Kari’s game into components that did have analogues in daily life: the concept of opportunistically taking something from someone else; the concept of prosocially standing up for someone else, and so on. Then, to decide on a course of action given these components, they would need to consult a whole set of what are known in the jargon as internal regulatory variables (Tooby, Cosmides, Sell, Lieberman, & Sznycer, 2008). Internal regulatory variables are running mental meters of some aspect of the environment or your own state. For example, social trust is an internal regulatory variable that tells me to what extent the behaviour of others is likely to be benign and reliable. When a stranger asks me for help in some complex way, I evaluate his demand with reference to my level of social trust. If my level of social trust is high, then I am prepared to believe his story and risk helping him; if past experiences have made my level of social trust low, then I will be more sceptical and disinclined to help.

5Viewed in this light, the question of how acculturation happens becomes partly the question of how, through our experiences, the levels of our internal regulatory variables are set. They will no doubt be affected by personal interactions (every experience of betrayal, for example, probably brings trust down a notch). They will also be affected by the stories recounted to us by others about their experiences. However, there is also an important role for immediate perceptual input, and it is this role I particularly wish to explore in this chapter.

Perceptual experience and context sensitivity

6Every time you go out of your house, you are exposed to a barrage of perceptions: where people are, what they are doing, how they look, the state of their houses, the state of public structures, and so forth. Even if you have no interaction of any consequence with these people and structures, they constitute a rich source of information about what might happen if you were suddenly called upon to interact with them. Even the simplest act of visual perception, like recognizing that an object is a teacup, is more than a question of passively receiving photons. It is a process of active inference from the patterns of photons received, in which the mind reconstructs aspects not directly present in the signal, such as the shapes of obscured parts of the objects. I suggest that these inferential processes go well beyond just deciding what kinds of objects are out there, and calibrate the internal regulatory variables governing our social attitudes too. That house has barbed wire on its gate (perception); the people living in it feel at risk from intrusion (inference); I should be careful living in this community (updating of internal regulatory variable). Such updating could happen without any explicit verbal instruction, and without my needing to actually be a victim of violent intrusion myself.

7This view of acculturation as involving the setting of internal regulatory variables through perceptually-based inference leads us to ask how long it needs to take. A classical view would be that acculturation takes (and lasts) a lifetime: children are socialized in particular ways, the effects of socialization are gradual and cumulative, and the result is a pattern that, by adulthood, is largely set in stone. There are certainly data that support this view, such as the classic findings that young men socialized in the US Southern states behaved differently from those socialized in the North even when both groups were currently living in the same city and attending the same university (Cohen, Nisbett, Bowdle, & Schwarz, 1996). However, it is unlikely that all internal regulatory variables have to be set in stone. At least some of them remain flexible through life and are constantly being notched up and down by the flow of perception. We can imagine a continuum of cases. At one end would be cases where by adulthood, the amount by which the variable could move up and down from hour to hour with further perceptual experience was very limited. Most of the variation in the variable would therefore be between people with different long-term cultural backgrounds, and the variation within people over time would be negligible. We will call variables at this end of the continuum trait-like. The variation in these trait-like variables would still be environmental rather than genetic in origin, but once people were adults there would be little scope for further fluctuation.

8At the other end of the continuum would be cases where the variable still had a lot of room to vary in response to changes in the ongoing stream of experience. Put the person in one situation, the variable goes straight up; move them to another situation, it comes straight down. We will call variables at this end of the continuum contextually sensitive. Here, most of the variation would be relatable to the person’s current situation rather than their longer-term history. This raises the question of whether the psychological differences underlying the different behavioural outcomes between Neighbourhoods A and B are more trait-like or more contextually sensitive. (In Sampson’s [2012] terminology, this equates to the question of whether we are dealing with developmental or situational neighbourhood effects.) This question really matters because of its implications for social interventions. If the differences are mainly trait-like, then moving adults from Neighbourhood B to Neighbourhood A would not make them any more trusting; as the saying goes, you can take the woman out of the West End, but not the West End out of the woman. On the other hand, if the differences are mostly contextually sensitive, then moving adults from B to A, or perhaps more relevantly improving the physical environment of B, could have marked real-time effects on trust and other social outcomes.

9There is evidence suggesting that many of the internal regulatory variables underlying social behaviour are particularly context-sensitive. Perhaps the strongest comes from Keizer, Lindenberg, and Steg’s (2008) experiments. To recall, in their experimental conditions, they sowed the environment with small cues as simple as graffiti on a wall, bicycles parked in violation of a rule or supermarket shopping carts that had not been returned to the proper place. These cues were purely perceptual; they had no direct consequence for the participant. They were also very subtle, but they all suggested in some way or other that people around here were routinely failing to be prosocial. The consequences of the experimental treatments were that the participants started to behave antisocially in lots of other ways, some of them more serious, such as stealing €5. Thus, the participants were not just copying the behaviour implied by the cues, but were using the cues to recalibrate some internal regulatory variable that then affected their decisions in other social domains. Most importantly for current purposes, the effects were dramatic: in real time, through perceptual means alone, the experimenters made the normally prosocial Dutch into a rather antisocial community.

10We therefore want to ask whether the differences in social behaviour between the residents of Neighbourhoods A and B reflect context-sensitive responses to immediate experience, or whether they are more trait-like. Some evidence for context-sensitivity comes from the results of Kari’s norms experiment in Neighbourhood B, as described in chapter 5. We also carried out another ambitious experiment to try to probe for context-sensitivity, and this is described in the next section.

An experiment with minibuses

11An ideal experiment to investigate context-sensitivity would involve taking a group of residents from Neighbourhood A and moving them to Neighbourhood B, directing another group in the opposite direction, and having two control groups that stay where they are. We could then track key psychological variables such as trust over time: to the extent to which trust is context-sensitive, the two moving groups should come to resemble their new neighbours rather than their old ones. How fast the change happened would tell you something about just how labile the context-sensitive variables were. Something slightly akin to this experiment was performed by one of the most ambitious US federal housing programs ever, Moving to Opportunity, which provided some poor families with vouchers allowing them to move to more affluent neighbourhoods, whilst some went to other poor neighbourhoods and some stayed where they were. The consequences for the participants’ health and behaviour were complex, but, overall, they provide some of the strongest causal evidence we have for the importance of neighbourhood effects (Kessler et al., 2014; Ludwig et al., 2012; Sciandra et al., 2013).

12We discussed such an experiment at length within the Tyneside Neighbourhoods Project, and concluded that we did not have a workable way of doing it. However, we did come up with a design for a study that shared some interesting features with such an experiment, but in a much smaller way. This was the work of Gillian Pepper, Kari, and myself, ably assisted by Ruth Jobling and students Bobbie-Jay Hasselby and Anna Wilson. What if, instead of transplanting people from A to B and vice versa, we recruited a third group of adults, residents of neither A or B, and randomly assigned them to spend some time in the environment of one or the other neighbourhood? Let us call these two experimental groups A-visitors and B-visitors. We could then measure some important variables such as trust in both groups. The null hypothesis would be that there should be no systematic differences between them. After all, the two groups were formed by random assignment. If, on the other hand, people are sensitive to neighbourhood context, then the A-visitors should be more highly trusting than the B-visitors, just as residents of A are more trusting than residents of B. The more similar the psychological measures were in A-visitors to A residents, and B-visitors to B residents, the more we would feel it plausible that the differences between residents of A and B were the outcome of immediate responses to context. Of course, we would still be well short of proving that if you moved people from B to A they would soon begin trusting like the other residents of A, but it would nonetheless be an intriguing result.

13Two things became clear about this experiment. The first was that we did not want our participants to know about the hypothesis, or even the nature of the study design. It would be too easy for them to offer us the obvious stereotypes about different parts of the city if they knew that was what we were asking about. So the experiment had to be somewhat surreptitious; we had to get our participants into the neighbourhoods on some plausible ulterior motive and then slip our measures in. The second thing that became clear is that our experiment would have to be audaciously short. You can’t make people go and spend three weeks in a neighbourhood without really telling them why. It was going to have to be not much more than a quick walk around. This seemed to stack the odds against our finding anything. Our participants’ experiences of the neighbourhoods they were assigned to were going to be so small and fleeting that it was hard to believe we would see any measurable effect. However, there is some precedent for very fast-acting effects of exposure to a cultural context. It is well established that, in viewing scenes, Westerners tend to focus most on foreground objects, and Japanese more on the background and relationships between objects in the scene. What is less well known is that just making American participants view 95 images of Japanese street scenes causes their perceptual style to shift substantially towards a Japanese one, whilst making Japanese participants view images of the US shifts them towards the American style (Miyamoto, Nisbett, & Masuda, 2006). The streets of the USA and Japan just look different. Thus, part of the difference between Americans and Japanese is not so much a matter of long socialization but the way the perceptual system is being driven by its immediate inputs.

14For our experimental design, what we settled on in the end was a case of killing two birds with one stone. We were always needing to deliver questionnaires to addresses in Neighbourhoods A and B, as part of the ongoing social surveying. We decided to recruit research volunteers who would help with these deliveries for an hour or two. This involved them showing up at a meeting point in the university, and being randomly shown into one of two minibuses. One minibus would go to Neighbourhood A and one to Neighbourhood B. Each participant would be given a list of addresses, a personalised street map, and a packet of questionnaires. The minibus would wait at a central point and the participant would return to it when they had finished (the 52 participants we recruited ended up taking 10–48 minutes to make their deliveries). As soon as they were back at the minibus, we would measure the psychological variables we were interested in. I found this delivery paradigm promising. When you are trying to find an address to deliver a questionnaire, you really look around you. You see the different streets and types and conditions of housing as you search for the required address; you get lost and have to back up; sometimes you ask a local. In other words, you are for that period alert and richly immersed in and attentive to all the cues about social life that the neighbourhood’s streets have to offer.

15As for the measures, the idea was to measure in the minibus riders some of the very same things we had responses on from the residents of the two neighbourhoods. We had measures of social trust and personal trust from Social Survey 2 (see chapter 3), so we measured these on the same scale in the minibus riders. We also decided to measure paranoia, the feeling that others intend to do us harm, using a standard questionnaire scale. To get some resident data on paranoia, we recruited an extra 65 householders who got a new Social Survey containing paranoia as well as social and personal trust; it was — with pleasing neatness — these questionnaires that our minibus riders were tasked with delivering.

16The results from the minibus study are summarised in Figure 6.1. Each panel represents a different psychological measure. Within each panel, the first pair of bars represents the scores for the residents of the two neighbourhoods, and the second, the scores for the visitors (the resident paranoia scores have been adjusted for age, sex and non-local origin, all factors known to affect paranoia). Taking social trust first, amongst the residents there is the expected neighbourhood difference, with social trust much higher in A than B. What is much more surprising is that difference is mirrored almost perfectly in the minibus riders: visitors to Neighbourhood A rated their social trust higher than those who had visited Neighbourhood B. Moreover, the trust ratings of visitors to Neighbourhood A were not significantly different from those of Neighbourhood A residents, but were significantly different from those of Neighbourhood B residents, whilst the ratings of visitors to Neighbourhood B were not significantly different from those of Neighbourhood B residents, but were significantly different from those of Neighbourhood A residents. In other words, minibus riders who had been in neighbourhoods for 10-48 minutes looked exactly like long-term residents of those neighbourhoods. It is important to stress that we did not ask them to imagine how much they thought they would trust if they were a resident of this neighbourhood, or to guess how much the people they had delivered surveys to trusted, or to think about trust right now in particular. We asked them how much they, as a matter of fact, trusted people they met for the first time.

Figure 6.1 Mean social trust (left panel), paranoia (centre panel), and personal trust (right panel) for residents of and visitors to each neighbourhood. Error bars represent one standard error. Data described more fully in Nettle, Pepper, Jobling, and Schroeder (2014). Image © Daniel Nettle, CC BY.

17The results for paranoia (centre panel) are rather similar to those for social trust. There is the expected neighbourhood difference amongst the residents, with paranoia higher in Neighbourhood B than A. This pattern is reflected in the minibus riders: visitors to B are more paranoid than visitors to A, and the paranoia scores of visitors to a neighbourhood are statistically indistinguishable from the paranoia scores of the people who live there. Again, the paranoia questionnaire did not ask participants about paranoid thoughts at this moment in particular, or the paranoia they would hypothetically feel if they lived here, but their own self-assessment of their thoughts concerning whether others in general were out to do them harm.

18The third trait, personal trust (the trust of people you know well), did not show the same pattern. As already mentioned in chapter 3, there was a neighbourhood difference amongst the residents on this variable. There was no significant difference between the two visitor groups. In a way, though, this is the exception that proves the rule. By putting people into the general social environment of one or other neighbourhood, we had manipulated the informational inputs to regulatory variables concerning what to do in interactions with people they did not know. We had not manipulated any information about the people they did know well, and so we should not have expected effects on internal regulatory variables concerning interactions with known others.

19What the results for social trust and paranoia suggest is that key psychological variables that differ between Neighbourhoods A and B are highly sensitive to context, and are notched up and down by recent perceptual experiences of the social environment. Therefore, the results imply, all you have to do to make someone think and feel like a Neighbourhood B resident is make them walk about a mile in the shoes of a Neighbourhood B resident around the streets that a Neighbourhood B resident would see every day. This was for me a stunning finding: I imagined that the low trust and higher paranoia of B residents was the cumulative effect of many years (generations perhaps) of acculturation into a world of economic and social uncertainty. The idea that you could reproduce essentially the whole of the Neighbourhood A/B trust difference in a group of experimental volunteers within one hour without doing anything in particular to them, just by having them be there, made me consider the neighbourhood differences in a whole new light. Of course, I am not claiming that all of acculturation is a simple matter of immediate context-sensitivity, or that all neighbourhood effects are contextual rather than developmental ones. That is clearly not the case; if I send you to the Arctic for an hour, you certainly will not become Inuit. However, perhaps the psychological differences we observe between human groups are more a result of sensitivity to immediate perceptual context than we have previously imagined.

20The results of this experiment should be relatively encouraging. If the differences between people from deprived neighbourhoods and people from affluent ones were mostly trait-like, then there would be no easy social interventions. Making the West End look cleaner, safer, and better cared for would not be expected to produce immediate gains in terms of greater trust and prosociality: any such gains might be a generation in coming. The experiments by Keizer and colleagues (2008) already suggest that this is not right, and our minibus study seems, in a different way, to confirm the picture. Simply changing the way the environment looks might have substantial impacts on people’s internal regulatory variables, and hence, perhaps, their social behaviour. We will return to the potential for such interventions in chapter 7. However, you would not be able to get very far in designing them until you had explored a follow-up question: what is it that people see in Neighbourhood B that lowers their trust and raises their paranoia?

The social diet

21Our results, and those of Miyamoto and colleagues (2006) on Japanese-American differences, bring to mind so-called visual diet effects in the perception literature (Rhodes, Jeffery, Watson, Clifford, & Nakayama, 2003; Webster, Kaping, Mizokami, & Duhamel, 2004). Visual diet experiments have shown that you can rapidly alter people’s assessment of reality through manipulating their exposure to visual inputs. For example, if you show people a series of slightly angry faces, their view of what a normal neutral face looks like becomes, in a very few minutes, re-centred towards the angry end of the spectrum. Moreover, non-angry faces start to look odd to them. These visual diet effects are found for basic parameters such as size, shape, and proportions as well as socially-laden ones such as facial expression.

22Visual diet effects are very strong and very reliable. What is odd about them is that they should exist at all: participants in these experiments have had many years of life experience of what facial expressions normally look like. It is hard to understand why the brain’s perceptual mechanisms should be designed to devalue all of that accumulated information about the world and give so much weight to just a few dozen recent exemplars. For whatever reason, though, this is how it seems to work. Internal social regulatory variables such as trust might work in a similar way, and here it is perhaps easier to see why an extreme sensitivity to the most recent inputs might be a good design. People presumably have always moved frequently from social grouping to social grouping. The right level of trust has thus not been something that could be set once and for all, but rather something that constantly needed notching up and down every time to the social context changed. Thus, perceptual mechanisms that scoured the current environment for cues to the right current level would always have been useful.

23If basic perceptual regulatory variables like expectations about the shapes of faces are set by the visual diet — the set of recent facial exemplars offered by the environment — then the analogous concept for social regulatory variables like trust is the social diet. The social diet is the set of perceptually-available cues about how people in the current context approach social behaviour. Let us think about how the social diet available in Neighbourhood B might differ from that in Neighbourhood A. There are actually a lot of ways in which it will be similar: they are both urban contexts, with buildings of similar vintage, similar numbers of people around, similar vehicles and advertising, similar street signs. However, there are some differences that our data allow us to clearly identify.

24A first difference is that the environment of Neighbourhood B contains more cues of prior antisocial behaviour by others, the most striking being litter. In Observational Dataset 1 (chapter 3), we saw someone drop litter every 30 minutes in B as against every 3 hours in A. That is absolutely nothing compared to the difference in litter on the ground. My experience of Neighbourhood B is that litter stays on the ground for many weeks or months, exacerbated by the fact that the system of rubbish collections is not used properly, with many receptacles filled with the wrong materials or at the wrong times, broken, or spilling over. Litter is not the only cue of prior antisocial behaviour, although I believe it to be possibly the most important single one. The iconic broken windows are very much in evidence in Neighbourhood B (Figure 6.2), along with broken bus shelters and other broken structures. It is not uncommon to see evidence of recent fires, presumably deliberately set. These cues say not only ‘people round here are prone to behave antisocially’, but also, equally importantly, ‘no-one round here has been prosocial enough to clear this up’, both powerful messages.

Figure 6.2 Empty house in Neighbourhood B. This house sports both broken windows and social defensive measures such as razor wire and steel shutters. Image © Daniel Nettle, CC BY.

25A second difference is that Neighbourhood B contains many more cues of social defensive measures (Figure 6.2). Many yard walls in B are topped with razor wire or, particularly brutal looking, broken glass set into concrete. Windows and doors are sometimes protected with thick metal bars. A doctor’s surgery on the boundary of B is literally fortified, with anti-climb razors and steel armour on the windows (Figure 6.3). This is almost exactly how police posts looked in Northern Ireland at the height of the Troubles. The irony of these defensive measures is that they are no doubt undertaken with the intention of providing a sense of security. Their effects, however, could be contrary to the intention. Seeing the lengths people round here feel they need to go to to protect themselves powerfully communicates that round here is not a very safe place. Thus, the defensive measures could have the aggregate effect of people in the community ending up feeling, or even perhaps being, less safe than they otherwise would be.

Figure 6.3 Doctor’s surgery on the boundary of Neighbourhood B. Defensive measures include a 2m spiked fence, spiked anti-climb rollers, and armoured doors and windows. Image © Daniel Nettle, CC BY.

26A third difference in the social diet of the two neighbourhoods concerns police presence and formal law enforcement. The local police, like many police forces, are highly aware of the spatial distribution of incidents requiring their attention, and also concerned with being seen to provide community safety. They patrol Neighbourhood B at a much higher rate than Neighbourhood A. During the gathering of Observational Dataset 1, I was passed by 23 police patrols (mostly in cars; Figure 6.4) in B and 4 patrols in A. Coupled to the patrolling is use of surveillance: much of the West End of Newcastle is fitted with street camera installations (Figure 6.5). The already brutal looking camera towers themselves have to be defended with metal spikes or barbed wire (presumably since, to slightly misuse a bit of Juvenal, nobody is watching the watchmen). No doubt the motivation of the patrols and surveillance is to make people feel safer: if something bad happens to you, we will see it. However, these measures too could have paradoxical effects: if the police have to put that much resource into fighting crime and disorder here, the unconscious inference could go, it must be a really bad environment. So it is an open question whether the presence of cues to law-enforcement effort is overall reassuring or alarming.

Figure 6.4 Street scene with police car, Neighbourhood B. Image © Daniel Nettle, CC BY.

Figure 6.5 Surveillance installation amongst houses, Neighbourhood B. Image © Daniel Nettle, CC BY.

27The fourth difference in the social diets between the two neighbourhoods is that already discussed at the end of chapter 5: the relative abundance of older people on the streets of A and their relative scarcity in B. Young adults (especially male ones) are the most dangerous category of human, and moreover the lack of older people might lead to the inference either than this is not a place where life is long, or that this is not a place the frail dare being in the public space. Neither of these inferences would be likely to promote trust or reduce paranoia.

28Jessica Hill ran a project to begin to investigate the impact of social diet cues (Hill, Pollet, & Nettle, 2014). She was interested in understanding the effects of two types of cue in particular, cues of disorder and cues of police presence, and also how these two types of cue might interact. By disorder she meant litter, broken windows, neglected buildings and so on. The prior literature was fairly unanimous that people notice and respond to cues of disorder, but very mixed on the subject of visible policing. Some studies had found that it increased feelings of safety, others that it exacerbated the perception of crime risk, and several others that people did not really notice police presence at all. Jessica created three experiments in which participants were exposed to a virtual neighbourhood. In the first experiment the neighbourhood was described verbally, and in the other two it was made using a slide show of real pictures of Neighbourhoods A and B. Each experiment had a factorial design with four conditions. In the first, the virtual neighbourhood featured high disorder and visible police; the second, high disorder and no police; the third, low disorder and visible police; and the fourth, low disorder and no police. Having experienced their virtual neighbourhood, participants were asked to rate it firstly how safe they thought that environment was, and secondly what the social capital of the people who lived there might be (through a series of questions such as whether the people in this neighbourhood could be trusted and whether they were willing to help one another).

29The results of the three experiments were remarkably consistent. People responded strongly to cues of disorder, feeling much less safe and perceiving much lower social capital where these cues were present. This is compatible with a wealth of social science research, described above and in earlier chapters, showing that even minor disorder is infectious and prone to undermining social capital and trust. Cues of police presence, by contrast, had no impact whatever in any of the studies. People did not seem to notice them at all. I suppose this has positive implications for the police in that their greatly increased presence in high-crime areas does not seem to exacerbate the fear of crime in those areas. However, it is hardly encouraging if they expected that their visible presence alone would suffice to reduce the fear of crime or bolster the feeling of community trust.

30Jessica’s study is only a first step, and this part of the Tyneside Neighbourhoods Project represents unfinished business. The kinds of social diet cues we respond to and the kinds of inferences we make from them are topics eminently amenable to experimental investigation. As well as simple slide shows, there are more sophisticated approaches we could use by employing video, eye-tracking, or even immersive virtual reality. These kinds of investigations, as well as being interesting for fundamental reasons, are surely of applied importance, since local governments and agencies have to decide how to allocate their fixed budgets. Surely one of the considerations in such decision-making should be the likely impact of different allocations on citizens’ internal regulatory variables such as trust, and hence, on their social behaviour.

Table des illustrations

Légende Figure 6.1 Mean social trust (left panel), paranoia (centre panel), and personal trust (right panel) for residents of and visitors to each neighbourhood. Error bars represent one standard error. Data described more fully in Nettle, Pepper, Jobling, and Schroeder (2014). Image © Daniel Nettle, CC BY.
Fichier image/jpeg, 76k
Légende Figure 6.2 Empty house in Neighbourhood B. This house sports both broken windows and social defensive measures such as razor wire and steel shutters. Image © Daniel Nettle, CC BY.
Fichier image/jpeg, 1,1M
Légende Figure 6.3 Doctor’s surgery on the boundary of Neighbourhood B. Defensive measures include a 2m spiked fence, spiked anti-climb rollers, and armoured doors and windows. Image © Daniel Nettle, CC BY.
Fichier image/jpeg, 782k
Légende Figure 6.4 Street scene with police car, Neighbourhood B. Image © Daniel Nettle, CC BY.
Fichier image/jpeg, 872k
Légende Figure 6.5 Surveillance installation amongst houses, Neighbourhood B. Image © Daniel Nettle, CC BY.
Fichier image/jpeg, 720k


Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search