© Daniel Nettle, CC BY http://dx.doi.org/10.11647/OBP.0084.01
It’s like a jungle sometimes. It makes me wonder,
How I keep from going under.1
This book is a study of people’s social relationships and social behaviour in different neighbourhoods of one English conurbation, Tyneside. I define social behaviour in the classic biological sense, to mean things that one individual does that have consequences for another individual or individuals (Bourke, 2011; Hamilton, 1964). Thus, merely being present at the same place as another individual is not necessarily a social behaviour. Giving something to them, taking something from them, improving their environment or despoiling it are all social behaviours. Social behaviours can be further classified as prosocial, where the actor’s actions improve someone else’s situation, or antisocial, where they make it worse. For the most part, the book is based on comparative data from two particular neighbourhoods within the city of Newcastle upon Tyne, which I will call A and B. There is also some ancillary data from other parts of Tyneside. The book is based on several years of intermittent fieldwork by several people; I will say more about this later in the chapter. First it is worth saying something about how the long journey that has led to this book got started.
In around 2007 or 2008, I had a chance conversation with my partner on the subject of household refuse. I lived in the North East of the city of Newcastle upon Tyne, whereas she lived in the West End, in the house we now live in together. Newcastle, like many cities, asks its citizens to divide household waste into recyclable and non-recyclable categories, collected on different days. I was uncertain, as I recall, about the proper categorization of some kind of plastic. I wanted to behave well and sort correctly, and was therefore anxious to have the right information about what the rules were. She said something to the effect that she wouldn’t expend too much effort on getting things like this right, since her bin would be quite likely to be set fire to anyway before it could be collected.
It turned out that her attitude was not without foundation. Various things did indeed get set ablaze in that part of Newcastle at that time, a problem that thankfully seems to have abated somewhat. Even when recycling bins did not burn, they would often be either kicked over or used improperly by someone else once on the street, resulting in a mixed load that the recycling lorry would refuse to take. Thus, she was absolutely correct in her assessment of the futility of expending much effort in the direction of conscientious recycling; such effort would end up being undermined by the action of others. Her lightly-made comment made a remarkable impression on me, for several reasons.
First, over the years I have thought a fair amount about the age-old question of whether people are basically good (helpful, prosocial, cooperative), or whether they are basically selfish. This question has a very clear answer: it depends. Humans have motivations to deliver social benefits to others, but these are not their only motivations, and the expression of these motivations is contingent and conditional. Most obviously, and as illustrated by the blazing bin example, the expression of prosocial motivations depends on expectations about what others in the population might (or might not) do. This means that if you want to understand when people will behave prosocially and when they will not, you need to know a lot about their ecology (what kinds of things are going on in the surrounding population?), and you also need to know a lot about human psychology (how exactly do the information-processing mechanisms that take cues from the local environment in order to adjust an individual’s social decisions work?).
The second reason that her comment struck me was that our respective houses were in the same city and only around 3km apart, yet clearly the behaviour going on around them was utterly different. Rubbish bins would never be set fire to where I lived. It was—and I mean this as more than a casual simile—like living in two different countries. Yet our two neighbourhoods shared the same language, ethnic heritage, national and local government, and judicial systems. In fact, the same council vehicles collected the refuse from the two places. This relates to the whole issue of the nature and scale of variation in human social behaviour, and in human culture more broadly. As social scientists, what should be our units of analysis: countries, cities, streets or individuals? How meaningful is it to talk about an English culture, when two samples of English people—two samples taken a 15-minute bicycle ride apart—give such different pictures?
Third, my partner had clearly and without much thought calibrated her actions to her ecology. She had a similar long-term developmental history to mine, and her fundamental social attitudes were the same as my own. Yet her neighbourhood environment had clearly caused her decision-making to change. Anthropologists call the process by which an individual’s social behaviour is shaped by the surrounding population acculturation. However, many descriptions of acculturation envisage a slow, perhaps linguistically-mediated process lasting many years, typically happening to children as they grow into adults. My partner had moved to the neighbourhood already adult; what had happened to her seemed more like an immediate cognitive response to a certain kind of perceptual experience. This raises interesting questions about which experiences are important in acculturation, and the timescale over which they act. If I moved to the land where bins go up in flames, would my behaviour change? If so, which perceptual inputs would be most important in causing the change, and how quickly would it occur if I had them? And if the change occurred quickly, how quickly could it be reversed?
If one tributary stream of this book was a longstanding interest in prosocial and antisocial behaviour, the second tributary was an interest in socioeconomic deprivation and its consequences. I haven’t told you, though it may not surprise you to learn, that the neighbourhood where recycling was overshadowed by arson was one where most people were extremely poor, whereas the orderly neighbourhood was one where most people were affluent. If there were large differences between our respective neighbourhoods in terms of prosocial and antisocial behaviour, we might be dealing with another instance of the near-ubiquitous phenomenon of the social gradient.
Social gradient is the term used by social scientists to describe any situation where the outcome we are interested in is patterned according to socioeconomic position, so that more affluent or high-status social groups look different from less affluent or lower-status ones. I may be coloured by the particular topics I have conducted research on, but social gradients strike me as the overwhelmingly salient fact about contemporary developed societies. If a Martian researcher asked me for a quick summary of how these societies work, I would give the following one: things work out differently for the rich and the poor. Social gradients have been described for many variables in the UK: birth weight, age at parenthood, paternal behaviour, breastfeeding, smoking, body mass, depression, and orientation towards the future, to name but a few (Adams & White, 2009; Nettle, 2008, 2010a; Pill, Peters, & Robling, 1995; Stansfeld & Head, 1998). Perhaps the most fundamental social gradient is that of life itself: the poor in the UK can expect to be alive several fewer years than the rich, and they can expect to be healthy for many fewer years (Adler, Boyce, & Chesney, 1994; Bajekal, 2005). Whether the gradient in the length of life is the cause or the consequence of all the other gradients is a delicate question. In my work, I have argued that there are often bidirectional relationships: poor people worry less about the long-term health consequences of smoking because they don’t think they will remain alive so long anyway, regardless of what they do, but this in turn exacerbates the already-existing gap in how long they will live (Nettle, 2010b; Pepper & Nettle, 2014).
Social gradients connect in a number of ways to the issues about prosocial and antisocial behaviour that I have already discussed. Some of the social gradients that have been documented directly concern prosocial and antisocial behaviour: there are social gradients in crime, in violence, and in pro-environmental attitudes, for example (Dietz, Stern, & Guagnano, 1998; Kikuchi & Desmond, 2015; Sampson, Raudenbush, & Earls, 1997; Shaw, Tunstall, & Dorling, 2005). Moreover, social gradients lead us once again to the question of the scale of variation in human culture. Because of social gradients, the variation within contemporary societies is often more striking than the variation between them. For example, Figure 1.1 plots women’s average age at first pregnancy for a number of countries, and then for two different groups of English women of White British ancestry: those who live in the most affluent decile of English neighbourhoods, and those who live in the most deprived decile. As you can see, the English women from the affluent neighbourhoods look like the average women from Switzerland or New Zealand. The English women from the deprived neighbourhoods behave like the average women of Guatemala or Kazakhstan. These groups of women often live just hundreds of metres apart, and yet we see that their lives are organized as differently from one another as Swiss women’s are from Kazakh women. This is immediately reminiscent of my short journey across Newcastle from the land of recycling to the land of burning bins.
Another connection between social gradients and prosocial behaviour comes from the fact that social gradients are not completely reducible to individual characteristics. Social scientists distinguish between an individual’s personal socioeconomic position, as measured by things like his income, educational qualifications and employment status, and the deprivation of the area in which he lives, which relates to the average income, education and employment of people in the surrounding locality. It is quite hard to tease apart statistically which is more important in social gradients, personal socioeconomic position (e.g. being poor oneself) or area deprivation (e.g. living in neighbourhood where many other people are poor). This is because the population of Britain is strongly assorted by income, so that most people living in neighbourhoods with many poor people are themselves poor. The best evidence suggests, though, that for many gradients, there is an effect of area-level deprivation above and beyond the effects of individual socioeconomic position (Ludwig et al., 2012; Pickett & Pearl, 2001; Sampson, Morenoff, & Gannon-Rowley, 2002). That is, there are consequences for one’s health and behaviour of living surrounded by poor people, above and beyond the consequences of being poor oneself. This must imply that our experience of what others in the immediately surrounding environment are doing is important for our own outcomes. This again links us back to social behaviour; indeed to the very definition of social behaviour as things people do that have effects on others.
My twin concerns with social behaviour and with socioeconomic deprivation were allied to a desire to get out of the office more. I had been doing epidemiological work for several years, and what this amounted to in practice was sitting in front of a computer. Although the scientific payoffs for desk work are often considerable, its capacity to expand personal horizons is limited: there is nothing quite like the messy improvisation of a primary empirical project for changing the way you think about the world. I thus decided—after considerable inspiration and advice from my friend Tom Dickins of Middlesex University—to undertake a systematic field study of two contrasting neighbourhoods within Newcastle, one very deprived and the other more affluent. The study aimed to be both ethnographic and ethnological. It would be ethnographic since it aimed to document, in detail, what life was like in the two study neighbourhoods. It would be ethnological, since I wanted to systematically compare the two, and try to explain why they might differ in the ways they did. I dubbed the project the Tyneside Neighbourhoods Project, partly as an homage to David Sloan Wilson and Dan O’Brien’s Binghamton Neighborhood Project in upstate New York (Wilson, O’Brien, & Sesma, 2009). The Tyneside Neighbourhoods Project ended up gathering data about many things, such as health behaviour, psychological wellbeing, and plans for the future. However, there was a core running through it that specifically concerned social relationships, social behaviour, and the cognition that underlies them. It is this core that forms the subject matter of this book.
I should say something at this point about what this book represents and how it is presented. The Tyneside Neighbourhoods Project started out as one study, but went on to spawn a series of linked follow-up investigations in the same sites. I carried out and wrote up the first study myself, but after that, I was fortunate enough to be able to recruit a series of wonderful collaborators. The ones whose work is discussed here were Agathe Colléony, Rebecca Coyne, Kari Britt Schroeder, Jessica Hill, Ruth Jobling and Gillian Pepper. Later parts of the project were executed and in many instances conceived by one or more of them, aided in some cases by undergraduate students from Newcastle University. I also collaborated on related work with Maria Cockerill of North Tyneside Council, and with Stephanie Clutterbuck and Jean Adams, though that work does not feature so directly here. All this led to a series of over ten papers in academic journals, each authored by a different combination of collaborators and each with a slightly different research question, but all based on fieldwork in Tyneside neighbourhoods.
I faced a difficult dilemma in the authorship of this book. The work presented here is not mine alone. Should I then co-author the book with all of my collaborators? Write it as an edited volume, with each chapter having a different author list? In the end I decided to author it by myself. My reasons for doing so relate to the function of the book. There is something interesting in having so many related datasets, using different methods, from the same places. It is like crossing and re-crossing a landscape from different angles, learning more about it each time. Now that the individual datasets are all published, I want to look at them in toto and reflect on the story they tell. This means looking at the whole body of the data from a little further back than is possible within a single academic paper.
Given that the function of the book is to stand back and take an overview, it seemed necessary to author it alone. An overview is necessarily someone’s view. Any of my collaborators might have come up with quite different overviews from my own, and it would be hard to corral those into a single coherent story. It would either end up as a book designed by a committee, or with my collaborators having to put their names to things they wouldn’t have expressed in the same way. Thus, I have written the book myself but on the basis of studies that were in many cases led by my collaborators. I will attempt to apportion credit (or, I suppose, blame) for the original studies by naming the key investigators and citing the original papers at appropriate points. Conversely, my collaborators are absolved from all the failings of the book; all opinions, errors, omissions and idiocies that appear in these pages rather than those of the peer-reviewed papers are mine alone.
There are some other things to say about the way this book is written. In my view, the main virtue of the project is the amount of quantitative data we were able to collect (see chapter 2 for more details). I have thus not been shy of presenting data in quite a lot of detail, especially visually (and for those wanting to delve deeper still, the raw data are available via Open Science Framework at https://osf.io/ys7g6).2 On the other hand, given that the objective is an overall synthesis, readability is paramount. There is nothing more inimical to easy reading than a lot of results from statistical modelling. I have therefore decided not to present any details of inferential statistics in this book. If I claim a difference or association is statistically significant, then this is backed up out of sight by appropriate statistical tests, usually to be found in the relevant published paper.
On a related note, the main text is very light on references to the previous literature, and quite heavy on my own interpretations and musings. I have eschewed an extensive literature review section in favour of mentioning key academic influences as they come up. Again, this relates to the different function of the book from the papers. Each of the published papers has a more thorough, better-referenced introduction, and a more measured and technical discussion. Here, I want to emphasize the story of doing the research as it happened, and I don’t want to get caught up in relatively esoteric points of difference between academic theories or traditions. I thus apologize to those scholars whose work I have been informed by but do not here cite explicitly, and also for the vast areas of social science literature of which I provide no adequate review in this book, even though it might be relevant. You can’t be nimble at the same time as being exhaustive, and so I have aimed for nimbleness.
In the remainder of this chapter, the next section will introduce the study city, Newcastle upon Tyne. The section after that says a little more about the academic concerns that shaped the project. As I have said, I don’t want to review these at length, but it is important to give some explanation for why the project ended up taking the form that it did. In the final section, I outline the broad hypotheses that we had in mind as we set out on the research.
The city context: Newcastle upon Tyne
Newcastle upon Tyne is the central city in the Tyneside conurbation, a major urban area of Northern England that occupies the banks of the river Tyne. The total population of Tyneside is around 900,000, making it the seventh largest population centre in the United Kingdom. Newcastle is an ancient city: the eponymous new castle was built in 1080 as part of the attempts of the Normans to secure—and harry—the North of England, but the city has been continuously inhabited since the Roman settlement of Pons Aelius begun in around AD120. Tyneside is a port. By the Middle Ages it was a major exporter of wool, and, increasingly and importantly, coal. Coal is found abundantly within and around the city, and coal was already being exported from Tyneside as early as 1250. By 1644, when Parliamentarian forces blockaded Newcastle during the English civil war, London’s annual coal supply fell from 450,000 tons to 3000.
Coal mining, as well as providing large-scale employment directly, drove industrialization of Tyneside, which during the nineteenth became a major centre for railway engineering, shipbuilding, armaments, and other manufacturing. The city expanded to the West, East and North, and its population swelled with immigrants from Ireland, Scotland, and other parts of England. Adjoining villages were incorporated into the city through the nineteenth and early twentieth centuries, as thousands of terraced homes were built in them for miners and other workers. They became the working-class neighbourhoods of Tyneside. (For a historical ethnography of life in one such neighbourhood in the early twentieth century, see Williamson [1982].)
The post-1945 period saw the beginning of economic decline and deindustrialisation on Tyneside. The last coal mine within the city closed in 1954; the larger mines in the surrounding area several decades later. The major heavy industries, notably shipbuilding, also declined through the 1970s and 1980s. Thus, the principal sources of employment that communities had grown up around employed many fewer people or disappeared. New sources of employment developed, notably in services, government-sector jobs and in education, and many parts of the city have regained and surpassed their former prosperity. However, this growth has been spatially uneven. As a result, architecturally and historically fairly similar neighbourhoods in different parts of the city have taken very divergent trajectories. This can be seen clearly by mapping the Index of Multiple Deprivation, the UK government’s preferred measure of social deprivation, for Newcastle’s constituent census tracts (Figure 1.2). As the map shows, the city contains many areas that are in the 50th-100th percentile when all of England’s census tracts are ranked from most to least deprived (that is, they are in the less deprived—or most affluent—half), but at the same time, also contains many tracts that are amongst the 10% most deprived. In fact, though the map does not show it, Tyneside has a number of tracts in England’s 1% most deprived. In particular, there are concentrations of extreme deprivation—the large nuclei of dark shading—along the riverside to both the West and East of the city centre. These were areas particularly tied to riverside heavy industries that no longer exist. By contrast, the area to the North East of the city centre, away from the river, although only a few kilometres distant from areas of extreme deprivation, is uniformly below median deprivation. The two main study sites for the Tyneside Neighbourhoods Project were one (Neighbourhood A) in the North East of the city, and one (Neighbourhood B) in the West End. I will deal with the choice of these two sites and their specific characteristics in more detail in chapter 2.
A very large housing-cost gulf developed between the North East of the city and the areas along the riverside. The deprived nuclei were characterized by a long-term pattern of underemployment, economic insecurity and physical dilapidation. This was accompanied by demographic loss; in the West End, for example, the population has declined by around one third over the last few decades (Robinson, 2005). The deprived areas developed negative reputations locally, fuelling a vicious cycle of further outmigration by those with economic options, increasing concentration of economic disadvantage, and a general sense of decline. In the words of one authority on the city’s development:
[The West End’s] reputation is legendary […]. One small, telling example of this reputation that I have experienced is of officials going to a meeting in the West End trying to avoid taking their own cars. Another example, illustrating the impacts of the image, is that people in the West End say that employers and others discriminate against them by association, simply because of their address or postcode […]. Discrimination and the area’s reputation encourage those who can do so to leave. Indeed, it has been disconcerting for those running regeneration projects to find that if people are helped to get jobs they are inclined to move out. It doesn’t help that the image seems to be starkly and graphically emphasised by the visual reality. Nowadays, there are parts of the West End that look appalling, with boarded up and burnt out houses, cleared sites, barbed wire and shuttered shops (Robinson, 2005).
The problems of deindustrialisation were further exacerbated by mortgage lenders applying no-lending policies to the deprived parts of the city. This further cut the already low value of housing, and restricted the pool of potential residents further still. It was essentially a signal of no confidence in these communities continuing to exist at all.
The deprived parts of the city have been the site of many different urban regeneration initiatives funded by local and national government (non-exhaustive list for the West End; 1960s: Urban Aid; 1970s: Benwell Community Development Programme; 1980s: Tyneside Enterprise Zone; 1990s: West End City Challenge, North Benwell New Beginnings, Scotswood Regeneration, Reviving the Heart of the West End, New Deal for Newcastle West). The most recent initiatives in the West End stem from the 1999 Going for Growth plan, and have involved large-scale demolition of housing, with a view to creating communities afresh, attracting people not currently willing to live there. The fieldwork described in this book was carried out in the period after much of the large-scale demolition envisaged by Going for Growth had been carried out, but before much of the new housing had been built. The West End made for an eerie sight through this period. In parts of Benwell, streets with street lighting were still there, but they had no houses on them. The former neighbourhood of Scotswood was a huge fallow field, with rocks blocking the ways in and, tall, angular CCTV-camera towers poking through the long, poppy-filled grass. Even the streets had been taken up. The Scotswood site is now (2015) being filled with the first houses of the new neighbourhood messianically named The Rise. Elsewhere, new housing is complete and already being occupied.
Neighbourhood B of this study consists of streets in the West End that were not demolished, and were not due to be. They were still fully functional and the houses largely inhabited. However, it should be borne in mind that the fieldwork was carried out at a particular moment in time: a moment when people in Neighbourhood B had endured many years of uncertainty about the future of the whole area, had seen many areas around them destroyed, and had not as yet seen much new put into the empty spaces that surrounded them. It is impossible to say, though interesting to ask, how different our results would have been had our study been done five years later, just as it is interesting to wonder what they would have been like if it had been done fifty or seventy years earlier, when the West End was still poor, but the basis of its economic life was very different.
Motivations for the Tyneside Neighbourhoods Project
The idea of studying in detail the social life of one or more urban neighbourhoods is by no means a new one. The project fits in some respects rather closely into the empirical tradition known as community studies in sociology (for a review of this area, see Crow & Allan, 1994). Community studies have often focused on working-class neighbourhoods, including on Tyneside in some cases, and often compared multiple sites in either the same or different cities (see e.g. Byrne, 1989). The present project shared the goals of some classic community studies, in that we wished to systematically describe, in our study neighbourhoods, the informal social relationships and interactions that make up so much of everyday life. (Strictly speaking, it would be better to call it a location study rather than a community study, since the study units were defined geographically and we cannot assume a priori that they necessarily function as communities.) However, my background is not in sociology, and I admit I was poorly acquainted with the community studies tradition of research at the time of setting out—so much so that some of the similarities of the Tyneside Neighbourhoods Project to the studies in that tradition represent lucky convergence rather than inheritance. The idea of studying different neighbourhoods within the same city nonetheless attracted me, because it seemed to relate to several broader theoretical issues that cut right across the social and behavioural sciences. The first of these is one I have already touched upon: the issue of scale of variation in human social behaviour.
Through the 2000s, a series of studies had shown quantitatively that measures of social behaviour varied across cultures (Henrich et al., 2005; Henrich, Ensminger, McElreath, & Barr, 2010; Herrmann, Thöni, & Gächter, 2008). ‘Cultures’ here almost always meant either countries or ethnic groups, and the measures were either surveys or the monetary dilemmas known as economic games that we will discuss in more detail in chapter 2. These studies were important in showing that you could not make simple generalizations about whether humans were or were not cooperative; you needed to specify which humans you were talking about. In a way, though, these studies were not the most striking demonstration possible of the extraordinary flexibility of human behaviour. The kinds of groups compared in this body of research differed in so many ways—their natural environment, their mode of subsistence, their level of economic development, their monetary and political systems, their languages—that you really ought to expect them to differ on measures of social behaviour. It is surprising that they did not differ even more. The proportion of variation explained by country or ethnic group in these studies was fairly modest: though the group differences were statistically significant, there was always much more variation within groups than between (Bell, Richerson, & McElreath, 2009; Henrich et al., 2012). The ethnic group has long been a privileged level of analysis in anthropological thought, and there is lively discussion in the literature on the extent to which this is justified (Henrich et al., 2012; Lamba & Mace, 2011, Morin 2015). To sort out the relative magnitudes of within-society and between-society variation requires what is called a multi-level study, where you have data from many individuals from many social groups from many societies. We weren’t going to be able to do that by conducting an ethnographic study within one city. However, we would be able to quantify between-neighbourhood differences and compare their magnitude to those seen between ethnic groups in cross-cultural studies using similar methodologies.
Another reason that a neighbourhood study was attractive relates to the distinction in social science between macro and micro studies. Macro studies are of the kind we have already discussed: they look at group averages in outcome across broad entities like countries or regions, often relating the variation in those averages to variation in some group-level predictor variable, either historical or contemporaneous. These studies are important because they reveal consequential differences between human social groups. One of the striking results of macro studies in the field of social behaviour has been to show that cultural differences can be rather stable over the long term. For example, there are marked differences between different parts of the USA in terms of the use of interpersonal violence. People in some areas are more prone than people in others to use violence to defend personal reputation, the so-called culture of honour (Nisbett, 1993). The prevalence of culture-of-honour violence in a county today, at least within the Southern states, can be traced right back to that county’s pattern of lawlessness and violence in the eighteenth century (Grosjean, 2014). These historical differences still show an influence when you control for current economic factors, and although the homogenizing influences of contemporary US life have lessened them over time, they have not yet abolished them completely. This suggests that once certain cultural patterns are set in, they become self-sustaining, and can persist over many generations of individuals, and much change in the current ecology. Such a conclusion sits well with a view of humans as ‘cultural beings’, the view traditional in much of anthropology. Under this view, people help or harm each other in particular ways because they are acculturated in societies that have historical traditions of helping or marking each other in those ways. These cultural traditions have their own inertia and dynamics that decouple them at least to some extent from subsequent changes in the ecology. (I am not claiming, by the way, that all macro studies necessarily lead us to view humans as cultural beings, only that some macro studies produce data easily interpreted from this perspective.)
At the opposite extreme from macro studies are micro studies. Here, the units of analysis are individual people, and researchers explore, often experimentally, the immediate causal factors that underlie their decisions. In the field of social behaviour, for example, there has been a lot of experimental work looking at what kinds of contexts produce cooperative behaviour and what kinds do not. Some of this work is done in the laboratory. In one famous set of studies, people were formed up into artificial social groups in which they were invited to contribute to a common pool resource that would benefit them all (Fehr & Gachter, 2000). In one experimental condition, the players had no possible way of sanctioning other group members who did not contribute. In this condition, predictably, contributions declined over time and the common pool resource ended up poorly funded. The experimenters changed the rules so that people had a way of sanctioning non-contributors. With this change in place, contributions to the common pool resource increased immediately, and remained high over time. In fact, people did not even need to directly experience being sanctioned to change their contributions; they saw that the possible consequences were different, and recalibrated their behaviour accordingly.
Other micro studies of social behaviour have been done in the field. Perhaps the most impressive example is experiments by Keizer, Lindenberg and Steg (2008). In city streets in the Netherlands, the researchers introduced, on some days, subtle cues of disorder into the environment. These cues included graffiti, or bicycles locked to a fence in violation of a rule. They showed that on days when disorder cues were present, people were much more likely to drop litter themselves, take a prohibited shortcut, or even pocket someone else’s €5, than on days when those cues were absent. The effects were large. The difference between denizens of the Dutch cities on a day when there is graffiti in the environment those same Dutch citizens on a day when there is no graffiti is at least as big as the largest between-population differences in social behaviour that have been observed in macro studies.
The strength of micro studies is that they provide insight into causal processes underlying individual behavioural decisions. Perhaps the most striking generalization arising from micro studies of social behaviour, and exemplified by Keizer et al.’s results, is that people’s social behaviour is enormously influenced by immediate context. You can really change what people do by even a small manipulation of their environment, and you can do this in real time. This sits well with the view of humans as ‘agentic beings’ that is characteristic of economics amongst other disciplines. Under this view, people are able to respond strategically to the incentives and opportunities of their local context: change the environment, and people’s behaviour will follow suit. There is an interesting tension between thinking of humans as cultural beings and thinking of them as agentic beings: on the one hand, the long-term stability of macro patterns, suggesting social culture has its own independent transmission and inertia; on the other hand, the strategic flexibility of individual people, suggesting that social behaviour is an immediate strategic response to current context.
In truth humans are beings that are both cultural and agentic, and one of the most interesting challenges in the behavioural sciences is constructing detailed explanations of particular behaviours that do justice to both components. Individuals are indeed flexible and strategic, as the micro-experiments demonstrate. They certainly don’t have a fixed, inherited cultural formula that they follow slavishly regardless of current context. On the other hand, there really are—at least sometimes—long-term cultural traditions (Morin 2015). These exist because people in populations influence one another. The behavioural output from one person forms part of the environmental input to which his neighbours may respond, creating the potential for a cycle of transmission. Under certain conditions, the aggregate behaviour can end up stable over long periods of time, even though all the individual people are capable of behaving flexibly and strategically.
The idea of a neighbourhoods project appealed because it was at a meso-scale, offering some of the opportunities of a macro study and some of those of a micro study. It shared with macro studies the possibility of identifying group-level average differences; particular neighbourhoods have reputations for particular social cultures, and these can last for many years. On the other hand, neighbourhoods are small enough and similar enough to one another that you can drill down deeply into exactly what individual perceptions or experiences are driving any neighbourhood differences in behaviour. It might even be possible in one or two instances to use the experimental method, deliberately manipulating the experiences or cues that people receive, to catch the workings of the decision-making processes underlying people’s prosocial and antisocial acts, and thus potentially understanding the proximate determinants of a neighbourhood’s social traditions. The meso-scale thus seemed a fruitful terrain on which to try to reconcile the cultural and agentic aspects of variation in social behaviour.
The Tyneside Neighbourhoods Project thus aimed to examine within-society variation, to work at the meso-scale, and to explore the roles of both individual decision-making and cultural transmission in maintaining patterns of behaviour. It had another aspiration, which was to use multiple quantitative methods, with equal emphasis on the multiple and on the quantitative. The reasons for this aspiration, and the methods it led us to develop, will be laid out in chapter 2.
Competing narratives:
Kropotkin versus the Mountain People
One of the reasons it seemed interesting to study social behaviour in an affluent and a deprived community is that there were two venerable traditions of thought available to us that led to exactly opposite expectations about what patterns we should see. These traditions are too diffuse to be called hypotheses; they are really alternative framing narratives about deprivation and social life that scholars and commentators are recurrently drawn to. I shall call them, for reasons that will become clear below, the Kropotkin narrative and the Mountain People narrative.
Piotr Kropotkin, the great nineteenth-century polymath and anarchist thinker, spent long periods travelling in Siberia and Manchuria. He observed the terrible harshness of the environment, and the consequent precariousness of life for the animals and plants. What impressed him, though, was how individuals cooperated as a strategy for survival. Whether it was the hunting packs of carnivores, the breeding colonies and flocks of birds, or the family groups of small mammals, Kropotkin argued that the way individuals coped with environmental difficulties was by developing mutually-beneficial social relationships. Kropotkin argued from the outset that this observation should be as applicable for understanding human societies as it was for animal ones.
An accurate reading of Kropotkin’s book Mutual Aid (1902) would not be that cooperative relationships emerge only in harsh environments. Rather, Kropotkin saw cooperative relationships as a ubiquitous component of how individuals manage to survive under all circumstances. However, harsh environments did allow you to see the importance of mutual aid particularly clearly. For this reason, his name has become perennially attached to the idea that harsh environments foster prosociality (Smaldino, Schank, & McElreath, 2013). Within a contemporary developed country like Britain, ‘harsher’ can be roughly equated to ‘more socioeconomically deprived’.
For our study, then, if we began with the Kropotkin narrative in mind, we would expect that it would be in our deprived study site where informal prosocial behaviours would be most strongly expressed. Individual financial resources are scarcer in such communities. The financial returns on labour tend to be much lower, even where formal employment is available, and consequently people’s ability to purchase market solutions to their problems is generally weak. They have to find ways of coping with this difficult constellation of circumstances, and under the Kropotkinian view, they will do this by turning to, and investing in, informal prosocial relationships within their neighbourhoods. Their social relationships will be particularly strong, and they will use their social networks for mounting cooperative solutions to their economic difficulties.
You can certainly find empirical evidence easily read into framework of the Kropotkin narrative. Young and Willmott’s famous community study Family and Kinship in East London (1957) depicted a working-class life of numerous social bonds beyond the household, neighbourliness, mutual aid and communal solidarity. Subsequent community studies of other working-class areas reinforced this basic view. More recently, in a series of psychological studies, Paul K. Piff and colleagues have found that young Americans of lower socioeconomic position tend to be more generous towards others, and more prosocially oriented, than their higher-socioeconomic-status peers (Côté, Piff, & Willer, 2013; Piff, Kraus, Côté, Cheng, & Keltner, 2010; Piff, 2014). This seems to be largely driven by their being more empathetic and compassionate, and more strongly endorsing egalitarian principles: we are all in this together.
However, the authority of some of this evidence is contested. Young and Willmott have been heavily criticized for relying on rather narrow, at times impressionistic, sources of data, and hence, perhaps, seeing what it suited them to see (Crow & Allan, 1994; Day, 2006). Follow-up studies in the very same localities saw much more conflict and isolation (Cornwell, 1984; Holme, 1985). Reliance on participants’ public accounts had produced a much more positive view of social life than the disturbing concerns people would raise when asked in private (Cornwell, 1984). This led to widespread scholarly questioning of the romantic Young-and-Wilmott depiction, as well as of the methods that generated it. The barbed comment that the community study is the sociologist’s substitute for the novel is widely reproduced (Crow & Allan, 1994 p. xii; Day, 2006 p. 29), and seems to have these portrayals of working-class life in mind. Note that the critiques of Young and Willmott do not necessarily apply to the more recent work of Piff and colleagues, whose methods are very different. Nonetheless, they do remind us that we should enter the enquiry with appropriate scepticism about the Kropotkin narrative as it applies to urban deprivation.
In the opposite corner from the Kropotkin narrative is the narrative I will associate with Colin Turnbull’s ethnographic monograph The Mountain People (Turnbull, 1972). The Mountain People was written following fieldwork amongst the Ik people of Northern Uganda carried out at a time in the 1960s when drought had led to very severe famine. Turnbull describes a situation where all prosocial norms had basically collapsed; people were, in Turnbull’s words ‘unfriendly, uncharitable, inhospitable and generally mean’ (p. 32). People were so desperate they could focus only on trying to survive themselves. The elderly died first; children were left to fend for themselves and died second; husbands fed themselves at the expense of their wives and vice versa; the dying were abandoned or stolen from. People actively avoided creating social relationships, because of the obligations they create and the obvious difficulties of having to deal with desperate individuals. Turnbull argued that under these circumstances, there was nothing in Ik villages that could be described as social structure, and no norms or rules of conduct that could be thought of as social. All that went on was the struggle of each person to survive: the Ik ‘place the individual good above all else and almost demand that each get away with as much as he can’ (p. 101). The pervasive interpersonal attitudes were mistrust and fear, and the closest of neighbours were mistrusted and feared the most (p. 134).
In invoking The Mountain People to give a name to a narrative, it is important to be clear about what I am and am not implying. Firstly, I am not implying that Turnbull’s ethnography was in fact a veridical representation of Ik social life at that time. The evidential basis of this work, too, has been criticized (Beidelman, 1973). Indeed, Bernd Heine reported a couple of decades later that some of the Ik were interested in the possibility of taking legal action against Turnbull for his misrepresentation of them (Heine, 1985). Second, I am not invoking or endorsing Turnbull’s view in the book that Ik society was irreparably broken and it would be better if it died out (p. 285). His proposal that the Ik should be broken up into small groups, forcibly relocated, and integrated into other communities round the rest of the country (pp. 283-84), is shockingly autocratic to modern sensibilities, though it is oddly reminiscent of occasional think-tank proposals for what to do with deprived inner-city areas in the North of England. Finally, I am not proposing too literal a similarity between the situation of the Ik and that of deprived parts of Newcastle today. Some similarity there is, since the shadow of hunger is not unknown in the West End of Newcastle. An emergency-assistance food bank opened there in 2013. According to its website, on the day before it closed for the Christmas break at the end of 2014, it provided food parcels to 1057 people. In September 2015, in an incident that could have come straight out of the pages of The Mountain People, it was broken into and all the food stored in it stolen. Nonetheless, the dissimilarities between the two contexts are obviously profound.
What I wish to extract from The Mountain People is a central intuition. This intuition is that deprivation undermines prosociality, because people are so preoccupied by just getting by that cooperating with others is something they can ill afford to think about. Note how the logic of the Mountain People narrative is an inversion of Kropotkin’s. In the Kropotkinian world, having good social relationships is so fundamental to survival that only bourgeois decadence would allow it to be forgotten; when things get tough, that is when you will see prosociality most clearly. By contrast, in the world of the Mountain People, the foremost and obligatory thing you need to do is to get by yourself. Investing in social relationships is something you will want to do once the more basic individual wants have been secured, like one of the higher-level needs in Maslow’s famous needs hierarchy. When times are hard, prosociality will be driven from the landscape. As Turnbull put it:
It seems that, far from being basic human qualities, [prosocial tendencies] are superficial luxuries we can afford in times of plenty […]. Given the situation in which the Ik found themselves […] man has not time for such luxuries, and a much more basic man appears, using much more basic survival tactics (Turnbull, 1972 p. 32).
For the present case, if we begin our enquiries with the Mountain People narrative in our minds, we will expect that in Neighbourhood B, where relative poverty and economic insecurity are widespread, people will be much more focused on individual subsistence, and hence prosocial behaviour will be less in evidence than in Neighbourhood A. You can find empirical support for this expectation, too. Haushofer (2013) marshalled an impressive amount of evidence from a dataset called the World Values Survey, comprising responses from over one hundred thousand respondents in 43 countries. He showed that within each country, there was a social gradient in trust, and in a measure of prosociality. (The prosociality measure asked about helping family members rather than people in general.) Trust and prosociality were highest in the richest decile of income, a little lower in the ninth decile, lower still in the eighth, and so on, with the poorest decile reporting themselves the least trusting and the least inclined to help. The gradients were strikingly steep and smooth. Interestingly, Haushofer found that the between-country differences in trust and prosociality were largely extensions of the individual-level patterns. That is, poorer countries were on average less trusting and less prosocial than richer ones, just as poorer individuals within a country were on average less trusting and less prosocial than richer ones.
Which narrative do we choose, the Kropotkin or the Mountain People? It is not necessarily an either/or choice. It is simplistic to reduce prosociality to a single axis and predict that there will be either ‘more’ or ‘less’ of it in one neighbourhood than another. A more nuanced expectation would be that prosocial behaviour will be present in both of our study neighbourhoods, but will take different forms, and perhaps encounter different obstacles. The two narratives are therefore only that: framing devices we can keep in our minds as we begin to examine, in chapters 3 and 4, what the datasets actually tell us.
1 The epigraphs to all of the chapters in this book are from ‘The Message’, by Grandmaster Flash and the Furious Five, written by Joel Edwards, Robin Barter and Robert Post (1983).
2 An archived version of the dataset, preserving the data in the form it was at the time of publication, is available at http://dx.doi.org/10.17605/OSF.IO/W9Z2P