The Social Complexity of Immigration and Diversity (SCID)
Description
Vision
Ours has been dubbed the ‘age of migration’ (Castles & Miller, 2003). Immigration is a major political issue, with increasing media coverage, the emergence of and anti-immigration political parties (e.g. the BNP) and rising anti-immigration sentiment (McLaren & Johnson, 2005; cf. Ford, 2008). The issue of migration sits centrally within the wider debate about social diversity – often focused on ethnic and religious diversity – and fears about its effects on social cohesion (Cantle, 2001). We are still, though, a long way from understanding these issues. They seem to rest on beliefs about shared values, national identity and ethnicity, but cannot be divorced from the effects of social class, education, economic competition and inequality, as well as the influences of geographical and social segregation, social structures and institutions.
SCID will apply novel agent-based simulation techniques to
understand the impact of immigration and diversity on social integration,
cohesion and inequality. In particular, a chain
of simulations – from complicated and descriptive up to abstract and tractable
– will be used to bridge the explanatory gap between micro- and macro-level
evidence, allowing for a more complete and coherent understanding of the social
complexity associated with immigration, social integration and diversity. These
simulations will be used to probe the interactions between otherwise
inextricable processes, including: changing and overlapping social networks;
the complex and more or less visible signals and attributes that make up a
person’s social identity; the dynamics between the individual’s view of
themselves and others’ perceptions of them; and the dissonances inherent in the
membership of multiple social groups.
This will require a new generation of social simulation models, as well
as novel techniques for developing, checking and applying these chains of models. Together these will allow for a greater range
of sociological and psychological evidence to be simultaneously brought to bear
on these important issues. Building on an established collaboration between
social and complexity scientists in
Background
Although mindful of the economic and cultural benefits of immigration, governments are concerned about problems of the social, economic and political integration of immigrants, and native citizens’ responses to an increasingly diverse society (Cantle, 2001; Denham, 2001). Substantial evidence exists in many western democracies of immigrant disadvantage in labour market opportunity, social mobility, educational achievement and political participation (Li & Heath, 2007; Politics REF??). Theoretical accounts for this range from assimilation to multicultural pluralism. Some (e.g. Alba & Nee, 2003; Park & Burgess, 1921) are essentially optimistic, suggesting that disadvantages will disappear with acculturation. Other theories (e.g. Gordon, 1964) argue that even if acculturation comes about disadvantage will persist until entry into the “cliques, clubs, and institutions of society” has occurred (ibid, p.71). Bloemraad et al. (2008) suggest that entry into political structures may be the key to eventual socio-economic integration. However, while ethnic identity might form the basis for political mobilisation (Fieldhouse & Cutts, 2008), such multicultural approaches to politics may be perceived as a threat by some.
Governments are concerned about how growth in social diversity may affect social cohesion, and building social capital is one of the favoured solutions to the problem (Commission on Integration and Cohesion, 2007). Social capital refers to both individual and collective resources (e.g. social networks and norms) which help societies achieve desirable ends. Whereas Coleman (1990) and Borduieu (1986) emphasised the private benefits of social capital, Putnam (1996, 2000) has argued that social capital is a public good. Flowing from this, much research has documented the many alleged benefits of social capital creation (Halpern, 2005). Others have found negative relationships between diversity and social capital, not just for minority groups but also for general populations (Alesina & Ferrera, 2000; Costa & Kahn, 2003; Putnam, 2007). However, the interpretation of these empirical relationships and their underlying foundations are not uncontroversial, criticised for being simplistic in their conceptions of diversity, for ignoring structural and class biases and for conceptual ambiguity (e.g. Portes, 1998; Skocpol, 1996). Moreover, recent research on tolerance and societal diversity (Ford, 2008) and diversity and trust (Putnam, 2007) throws up a paradox – it seems that increasing diversity is associated with greater tolerance (as in ‘contact’ theory), whilst neighbourhood diversity is associated with lower tolerance (as in ‘conflict’ theories).
To understand these problems properly one needs a simultaneous and co-dependent model of individual characteristics and behaviours on the one hand, and macro-social influences on the other, influences that may themselves be complex, emergent features of the micro-level individual interactions. The crucial interdependence of micro-macro relationships has long been recognised in social theory (e.g. Coleman, 1990), but rarely are the specific, individual-level interactions actually explicitly and realistically modelled. Existing modelling tools and data are not sufficient to do this. Statistical models can indicate connections between factors but, by their very nature, tend to “smooth out” the complexity. For this reason (and others) simulations has been developed which seeks to explicitly represent the intricacies of social interaction in so called “agent-based” models. Going beneath a broad analysis of social factors on each other to some of the complex interaction of individual actors (Macy and Willer, 2002).
Most of these simulations have not sought to directly represent any observations but are more abstract, seeking to capture an idea with which to understand what is observed (Edmonds 2001). Such models include abstract representations of social networks and their development (Watts & Strogatz 1998); tag-based models of spontaneous cooperative group formation (Riolo et al. 2001; Edmonds, Norling & Hales 2009); and the convergence and divergence of opinions within a group (Deffuant 2001). Some of these are amenable to some analytic treatment. Whilst these are productive in terms of ideas, and can allow for experimentation on the theory (Edmonds & Hales 2005), they are difficult to directly relate to the individual behaviour and macro-level sociological evidence.
Recently research thread has emerged based upon simulation models that are closer to computational descriptions of the phenomena they seek to capture, integrating the available evidence in a complex model that is quite specific to a particular context. Examples of these include (Alam et al. 2007; Schwarz & Ernst 2009). Such models can relate fairly directly to evidence but can themselves be hard to understand and are well beyond the limits of existing analytic techniques. They are also relatively slow to execute and so experiments upon them have to be limited. Thus any single model along the simple-descriptive (Edmonds and Moss 2005) spectrum of possibilities has shortcomings. Whilst simple models have been rightly criticised by social scientists as not being grounded in sociological evidence, the descriptive models have been rightly criticised by complexity scientists as lacking rigour.
Research Hypothesis and Objectives
SCID will develop new complexity science to address important issues in three related topic areas: the effects of immigration and diversity on i) social trust, ii) socio-political integration, and iii) socio-economic inequality. Underpinning all themes will be the core construct of social identity, used to operationalise the key notions of social similarity and difference. Our principle aims are as follows.
To develop novel complexity techniques. SCID will develop techniques of working at complementary levels of abstraction. Learning from the models at one level to inform the content of the next, so that the more abstract captures some of the important processes of the more detailed and complex. Also to allow many different kinds of expertise and evidence to be related together in a structured and explicit manner.
To develop new families of complexity models. SCID will develop new families of social simulation models, focused on currently unexplored interactions between different social mechanisms that are thought to be important in society at large. In this way complexity science will be enriched by exposure to new emergent phenomena from the real social world introducing new exemplars of co-adaptive processes, for example interaction-bias and social network.
To link micro and macro social theory. Using novel methods, SCID will show how some aspects of the target social phenomena can be explained as emergent phenomena arising from individual social interactions, bridging the gap between the micro-evidence about individual behaviour up to macro accounts of aggregate trends and patterns. In the past this has been done in a broad and analogical manner; we will complement such accounts with simulation-based accounts that show explicitly the interaction and mutual causation of individual and social phenomena.
To inform policy makers in the area of immigration, diversity and social cohesion. SCID will develop techniques that reveal areas of ignorance in the current evidence, indicating where decision-making can be carried out with more confidence and where more caution might be needed. It will help focus subsequent evidence gathering efforts. It will help clarify some of the interactions that are occurring in the social areas of concern and suggest “early-warning” indicators of the emerging social changes.
The research hypothesis of SCID is that these aims are achievable using “chains” of related models at a range of levels of abstraction from the data gathered by social scientists up to mathematical analyses via a series of new simulation models. In this way we aim to explicitly and more carefully stage the abstraction from data to theory in a traceable manner, bridging the worlds of rigour and relevance.
In pursuance of these aims SCID will achieve a number of concrete objectives within each of three, closely related areas of social science. Thus there will be the following objectives for each of the focus themes (described below).
1. To collect, organise and prepare a catalogue of existing relevant micro and macro evidence/data on the relevant individual characteristics, behaviours and social processes
2. To collect, organise and critique existing formal and simulation approaches that may be relevant
3. To develop and validate a series of detailed, descriptive simulations that integrate as much of the available social science evidence as possible
4. To actively collect new data that addresses the questions thrown up by the assessment of descriptive simulations
5. To analyse (and check) the descriptive simulations by simplifying them using techniques from physics
6. To develop new complexity tools and approaches to integrate rigorous simple models with richer descriptive ones
7. To produce new hypotheses concerning immigration and diversity, challenging simplistic conceptions
8. To produce simulation-mediated assessments of policy alternatives and issues
This will not be a neat, linear progression, since many of these objectives are related and co-dependent. These loops and dependencies as well as how they will be managed will be described in the next section.
As well as this “vertical” sequence of objectives there will be cross-cutting cross-fertilisation between the target social themes. Although the research in each of these areas will develop somewhat independently, since it is important not to seek to generalise too quickly treating each sub-problem individually, the approach of SCID will mean that there is considerable intellectual cross-over between them. Thus there will also be “horizontal” objectives.
9. To demonstration how a multi-layered simulation approach can assess and inform social science theory
10. To assess and codify the SCID methodology of multiple-level modelling, its advantages, methods and difficulties, with a particular view to its further development and dissemination
11. To summarise and disseminate simulation structures and techniques that prove to be generally useful
12. To cautiously generalise from the theory and simulations concerning the different sub-areas as to what more general patterns and explanations are apparent
13. To influence the debate on policy issues with new perspectives and insights
Programme and Methodology
The basic method of the project combines the collection of evidence and the construction of simulation/mathematical models, with the results being what can be concluded when these two are analysed and confronted. The kind of models will range in abstraction from statistical models of data, up through detailed and specific agent-based simulations, abstracted individual-based simulation models, up to analytical mathematical models. The process will progress in bottom-up route starting at the concrete and ending at the abstract; a “staged modelling” approach, where the increasingly abstract stages will provide checks and understanding of the levels below. At each stage if models are inadequate or new issues come to light, lower stages will be revisited in an iterative fashion. Assessment and direction of all models will be collectively assessed, including, crucially, input from social policy advisors.
Figure 1: Integrating micro and macro social theory and data using simulation methods, revealing some of the intricate possible interactions that may lie behind more discursive explanatory accounts
As Figure 1 indicates, while there exists a wealth of data on social phenomena at the micro (individual) and macro (social) levels, social theory is rarely explicit in describing how these levels are integrated (e.g. Coleman, 1990). Simulation methods can be used to bridge this micro-macro gap, revealing and exploring the causal threads that result in emergent phenomena. It is thus complementary to existing social science – producing a range of possible explicit explanations in terms of intricate interactions and processes in contrast to semantically rich accounts that are certain to be somewhat correct but necessarily imprecise. The big challenge for SCID is the large “gap” between the abstractions of complexity science and the complications and richness of social phenomena – it is this gap that SCID aims to address. Thus SCID aims to achieve both rigour and relevance. It will do this in a number of ways, principally:
· Instead of single models SCID will develop “chains” of explicitly related models at different levels of abstraction, bridging the gap from evidence up to theory in a series of smaller, less ambitious jumps.
· It will develop new families of analytic and simulation model that include more of the complexity and interactions that are known to occur but whose importance is difficult to assess.
The descriptive models at one end of the chair are developed with micro-level evidence and validated against macro-level evidence (Moss and Edmonds 2005). The data-integration models are then modelled by more abstract simulation models that seek to test what is essential and not, thus allowing some more general understanding of them to be developed. Where possible the abstract simulation models will be analysed using mathematical techniques. Such “chains” of models, each model being validated against the next one will allow the real integration of complexity and social science. This relates the relevance at the concrete end with rigour at the abstract end in an explicit and staged manner, at the cost of increased effort in keeping all these levels “aligned” (Axtel et al. 1996). Thus the integration of complexity and social science lies at the heart of SCID and would be its main contribution. It achievement would be one of significant consequence for the explication of policy-relevant social complexity.
The overall order of activities is show in Figure 2 below. This is the approximate flow of work that will occur for each of the focus themes of social trust, socio-economic integration and socio-political integration described below. The start of the flow for each such area will be slightly staggered in time to allow a more even distribution of evidence and time for the lessons learnt from one area to be applied in the next (see Workplan). Each of these activities is described in a brief paragraph below, before the specifics for each theme are specified.
Figure 2. An illustration of the workflow of SCID – this will happen for EACH of the thematic areas, but staggered. The ovals are milestones and the little “suns” indicate dissemination points. The thicker arrows indicate the main iterative loops. ISC, TPG and CPM indicate the areas of responsibility. D Simulation = Descriptive Simulation.
Survey and collation of evidence at the micro- and macro-levels. Qualitative and quantitative evidence from the sociological, social-anthropological, political science, human geographical, social-psychological and behavioural economics fields will be reviewed and collated. These will provide a detailed description of the processes of perception, value and attitude formation, and behaviour, at the level of individuals and interactions. Data on macro trends in social trust, employment and political behaviour, along with demographic information, will be collated from national surveys (see subsection below on data collection), along with census data (managed in part by ISC) and Office for National Statistics data (for geographical and contextual information). ISC researchers will conduct these activities. Outputs will comprise of: a report on micro-level behavioural processes, and harmonised sets of survey and census data on social trust, employment and political participation in relation to migration, ethnicity, geography and other demographic variables. [Fieldhouse, Li, Shryane, PDRA 3 & PDRA 4]
Survey and analysis of existing formal models. There are a number of existing models of social phenomena which have been constructed by complexity theorists, and others, which are relatively simple, seeking to describe specific social effects with the minimum of input. Frequently these involve processes occurring on social networks, the theory of which has rapidly grown in the last decade or so, largely driven by theoretical physicists and researchers in related areas (Dorogovtsev and Mendes 2003, Newman et al, 2006). A survey of existing models of this type will be carried out; they will be classified and related to each other where possible. This will act as training for PDRA 2 and help prepare for when the abstract models are extracted from agent based models later in the project. [McKane & PDRA 2]
Descriptive
Descriptive Simulation Criticism
and Validation. These complex simulations will be
cross-validated in as many ways as the available evidence allows (Moss and
Edmonds 2005) but including using: aggregate statistics and the opinion of
experts and advisors. Such validation is
always partial, but reveals the strengths and weaknesses of the
simulations. Output is an assessment of
the simulations’ validity. [Li, Shryane,
Responsive Data Collection. The process of data review and collation and descriptive simulation building will reveal gaps in the availability of data with which to specify and constrain the simulation and analytical models. SCID will undertake primary data collection approximately halfway into the project to remedy this gap. It is anticipated that the major shortfall will be in detailed information on real-world social networks. Three junior researchers will collect data for up to one year from overlapping social networks of immigrants and non-immigrants in Manchester using snowball sampling methods, aiming for a sample of a least 300 inter-related individuals. Outputs will be a report on the gaps in data with which to test current theory, and a social network dataset. [Crossley Shryane, PDRAs 3-4, RAs 1-3]
Descriptive Simulation Analysis. Although relatively slow to
run, the descriptive simulations are open to inspection at all levels and
stages, a limited number of exploratory runs and focused simulation experiments
(
Abstract Model Validation. The abstract models will be
validated by comparison with the outcomes of the descriptive simulations. There
may be broad agreement between the two approaches, in which case further
investigations will seek to understand the differences in the outcomes. More
likely, the predictions will differ considerably, in which case some further
structure will be incorporated into the abstract model, which is judged likely
to bring it nearer to the descriptive simulation. More generally, a chain of
models will link the abstract model with the descriptive simulations, allowing
insights gained in any one model to be used to gain further understanding of
the others. [
Analytic Development. The abstract models will be formulated through an individual-based reaction dynamics. These will be stochastic models with the individuals interacting through a social network or in a spatial context. These individuals will typically have very few characteristics; adding more characteristics would eventually yield the models used in the descriptive simulations. Abstract models of the kind above are frequently amenable to analysis using the tools developed in stochastic nonlinear dynamics (van Kampen 2007) and the theory of dynamical systems (Strogatz 1994). We will use these methods in conjunction with numerical simulations to understand the nature and predictions of a broad class of models suggested by the work carried out in this project. [McKane & PDRA 2]
Social Theory Development. In SCID, hypotheses concerning each theme will emerge from the comparison of two new sources: the insights gained from analysis of the descriptive simulations and the more general but abstract indications from the formal analyses. These hypotheses may suggest new theory, but will be assessed critically from the context of the current understanding and thought gained from the available collected evidence. [All]
Application to Policy Issues. Staff from CPM and ISC will
review the policy relevance of the modelling results in consultation with the
advisory group and more widely at the annual workshops. To aid this process ISC
will produce a document reviewing current
Cross Theme Comparison and Generalisation. Once the research on all three themes has progressed to a suitably mature stage, the hypotheses and insights gained from these will be gathered and cautious generalisations made. [Crossley, Fieldhouse, Li, Shryane]
In order to focus the efforts of SCID and ensure it is firmly connected to specific evidence and relevant policy issues, it will (at least initially) focus on a number of more specific themes within the general one of Immigration and Ethnic Identity. However the project, as it progresses will be responsive to the guidance of the advisory board and the developing results from the project. These themes are now described.
Diversity, homophily and social trust
One of the concerns about immigration and the increasing
diversity of society is the effect it has on social cohesion. Over the last
decade, particularly since the social unrest in several
· Define a number of explicit measures of cohesion, diversity, homophily and social trust that capture some of their characteristics in a more formal way
· Assess the ways in which changing patterns of individual characteristics in the population (e.g. immigration) can affect social trust
· Assess how societal constraints on individuals affect social cohesion
We will develop models of networks that show realistic levels of bridging and bonding social capital – ‘bonding’ capital is inward looking, bringing together people who are alike, thereby reinforcing their exclusive identities, whilst ‘bridging’ capital refers to ties which link people with different social attributes. Homophily (‘birds of a feather flock together’) and group identity will be emergent features in these models, based on common rules of affiliation acting upon variations the micro-attributes of individual agents (based upon social-psychological evidence and models).
These models will be a new family of simulation models that represent both the close social network as well as a “tag-based” mechanism for markers of group-likeness. The explicit (but dynamic) social network will indicate those close enough that information or favours might be sought whilst the tags are a fallible indication of who might be in the same ethnic or class groups (and hence bias interaction choice with strangers). In these some of the attributes will be highly ‘visible’ to other agents in the simulation (e.g. dress, skin colour), others less so (e.g. education). Agents will vary in the extent to which they see attributes as desirable and important (i.e. that influence affiliation positively/negatively, strongly/weakly). These models will therefore extend the ideas of social distance discussed by Akerlof (1997). The descriptive models will be specified in such a way to allow experimentation with differing: birth rates, class structures and behavioural rules that bias the choice of who to interact with (e.g. those who seem of the same group). The models will be later expanded to include spatial distribution of agents so that geographical clustering can be also represented.
These simulation models suggest a new class of more abstract models based on the idea of social tags. The techniques, which use reaction dynamics alluded to earlier, will be applied to these models. Master equations (continuous time Markov chains) will be used to make analytic predictions, which will be compared to results found using Gillespie-like simulation algorithms. Previous work by McKane and co-authors show such techniques are very powerful, with several new results being obtained for systems previously thought not to be amenable to analysis.
This set of models will be used to assess the effects of introducing new populations of agents (i.e. ‘immigrants’) into established networks; how differing distributions of attributes and initial social networks in the immigrant and host populations may alter diversity and social cohesion; what the effects of possible biases for the selection of strangers to interact with might have; and what the main factors that appear to influence the development of social trust are.
Socio-political integration
Different ethnic groups are observed to have different propensities for political engagement (e.g. voting), even when allowing for other relevant differences (e.g. age, social class). Factors influencing this include mobilisation through social networks, levels of bridging and linking social capital among others. Shared ethnicity may motivate minority participation in a political process (Fieldhouse & Cutts, 2008) but also be perceived as threatening by the majority (Bloemraad et al. 2008). Politicians, seeking to set the agenda to their advantage might (deliberately or otherwise) act to increase the perceived differences between ethnic groups. The theme-specific research objectives are:
· Assessing how the distribution of political status and power might explain differences in degree and types of political engagement within different ethnic groupings
· Exploring the possible impacts of political manoeuvring on social cohesion
The models made to analyse this theme will focus on power and status (or class) shifts that may result from the interaction of social networks and ethnicity with those of a political process. To do this we shall expand the models above to include individual attributes representing outcome measures for political participation and integration (e.g. electoral registration and voting). As well as modelling individual-level attributes and processes that affect outcomes we shall also include top-down influences, for example the local and national activities of political parties (e.g. policy, selection of candidates from minority groups). These influences may be applied globally or, more realistically, vary in their impact depending upon individual level attributes (e.g. ethnicity, class, education). In addition to network structures that capture bonding and bridging social capital, we shall seek to model ‘linking’ capital (i.e. social ties to people in positions of authority) by including attributes that represent hierarchies of political power and influence (such as community leaders, councillors, MPs).
The individual’s attributes will include multi-dimensional representations of their political preferences. These play a mutable part as group membership markers, but are also the subject of political debate. Parties will be introduced as adaptive entities which seek to place their policies (fuzzy regions in this space) as well as focus issues (divisions of the space) so as to seek votes. An initial voter rule to be assessed is that voters choose the party nearest to them, but on the same side as the dividing issues they consider important. Thus in these models there will be the co-adaption of the parties’ policies and the social power networks. This goes well beyond existing spatial voter models, in that it focuses on the co-adaption between parties, issues and ethnically relevant social markers.
Abstract social network models where the status of individuals is reflected in the nature of the interactions between them have been constructed by McKane and collaborators in the context of usage-based models of language change (Baxter et al 2006). These ideas will be modified and extended to facilitate the construction of abstract models which emulate the descriptive simulation models. Analytical techniques to study such models are currently being developed, and the work proposed in this programme will give a considerable impetus to this activity.
We will explicitly test the multicultural ‘group threat’ hypothesis (Bloemraad et al., 2008), which suggests that tolerance of diversity may decline in areas where minority groups are politically mobilised along ethnic lines. We will evaluate the extent to which ethnic attributes in such situations would actually be proxies for class and economic differences. We will also evaluate whether similar processes may account for the stability of political groups in the majority population based around conceptions of ethnic identity (e.g. the BNP), where the dividing issues sought by the party are exactly the ethnic social markers.
Socio-economic inequality
The final theme concerns access to suitable employment. Research has shown that different immigrant and ethnic groups have varying access to and success in the labour market (Li & Heath, 2008). This may be due to lack of fluency in English, home-country educational qualifications being devalued, disruption in social networks, or outright discrimination. In this theme there are the following objectives:
· Assess the impact of the structure of social networks on employment levels in different ethnic groupings
· Identify what the main barriers to new entrants to the employment market might be
Recent migrants may have restricted social networks, based around family or ethnic connections, whereas those in the host population may have a range of networks to exploit, including location-based, work-based, hobby-based and those based around their children’s school. Thus the way in which different social networks combine and overlay can be crucial in terms of an individual’s access to information about employment and education opportunities. The barriers to developing these networks, particularly bridging and linking capital, may vary by ethnicity for migrant groups. Also the willingness of individuals to apply for jobs perceived to be beyond their status as well as the class and ethnic prejudices of employers may mean that an opportunity that is appropriate in terms of skill is not taken.
The context, the social advantages and disadvantages (SAD) of different groups, will be represented as individual attributes and network properties (e.g. that represent levels of bridging, bonding and linking capital). In modelling terms, we will explicitly represent multiple dynamic social networks (or equivalently a social network with different types of links) so as to be able to examine the interactions between them. We will provide agents with contrasting interaction rules so as to assess the plausibility of the result, initially the case whereby they adopt group norms compared to the case where they react adaptively (and rationally) given the constraints of their local social context.
As for the models of socio-political integration, explicit top-down influences will be included, for example, representing different levels of discrimination by employers in the labour market depending upon similarities in the similarities in terms of class-based and ethnic social markers.
Social networks with several different kinds of links have received comparatively little attention in the context of abstract models. The development of descriptive simulation models of this type will provide a unique opportunity to develop analogous abstract models which are anchored to realistic models of social change. The abstract models will be formulated, analysed and interpreted using the techniques of stochastic nonlinear dynamics (McKane & PRDA 2).
We will seek to test Rational Action Theory (Goldthorpe, 2007) predictions, by having agents with differing characteristics (e.g. class, ethnicity, migration status) acting according to the same adaptive rules in the labour market. We will assess the impact of the relative poverty of the social networks available to immigrants on their employment and subsequent social embedding. We will look at the patterns of the match of skill and job within given a range of social network structures and behavioural rules.
Data sources
There are many data sources available for characterising macro-level processes and outcomes. Indicative examples include the following. Socio-economic inequality. Li and colleagues pooled key variables from 5 million records of the General Household Survey (GHS) and the Labour Force Survey (LFS) from 1972 to 2005 (and during the project data till 2015 will be available). Fieldhouse and colleagues administer Samples of Anonymised Records (SARs) from the 1991 and 2001 (and possibly the 2011) Censuses with millions of records. Altogether there will be around 10 million records spanning over 40 years to track ethnic socio-economic trajectories. Socio-Political integration. BES, BEPS, Voting records (in conjunction with census data), British Social Attitudes (BSA) surveys. Three waves of the ESS. Social Trust and diversity. The British Household Panel Study (BHPS, with around 10,000 records per wave, started in 1991. The Citizenship Survey (N=15,000) started in 2001. Among other additional studies are the Cultural Capital and Social Exclusion (CCSE) survey of 2004 and the English Longitudinal Study of Ageing. Many of the surveys above (e.g. BSA, ESS) contain data on social identity. Other surveys from diverse fields e.g. Ethnic Minority Psychiatric Illness Rates In the Community (EMPIRIC), Fourth National Survey of Ethnic Minorities, contain information on the relationship between ethnic/migration identity, social capital and integration.
Relevance to Academic Beneficiaries
We anticipate diverse but significant results from this project. First, in the integration of complexity and social science itself – a set of exemplar projects demonstrating that this is feasible; ways of developing, checking and updating chains of models to enable the bridge between the two fields; and a set of methods to help guide those that follow. This is a general result that could be useful in a wide range of fields. Second, three new families of social simulation model will be developed and explored pushing the boundaries of complexity and laying open some of their dynamics. The results regarding social dynamics and the micro-macro interplay will be important in a wide range of social phenomenon. Third, we will develop a set of formal analyses of aspects of these simulations, making them more rigorous. Fourth, we expect that the programme will make a real contribution to social theory, helping to make concrete issues that are often underspecified and producing new and revealing questions for further study and analysis, where the simulations have revealed previously unsuspected complexities. The results regarding social capital and immigration should give new insights into these areas of social science. Fifth, it will throw new light on policy-relevant issues, in particular providing new measures and watch points that may indicate newly emerging trends, as well as providing test-beds for assessing some of the effects (intended and otherwise) of possible policy interventions.
References (in addition to those in the Track Record)
Alba, R., & Nee, V. (2003). Remaking the mainstream: Assimilation and contemporary immigration. Harvard UP.
Alesina, A., & La
Axtell, R. et al. (1996) Computational and Math. Organization Theory 1(2):123-141.
Bloemraad, I. et al. (2008). Annual Review of Sociology, 34: 153-179
Bourdieu, P. (1986). The Forms of Capital. In Handbook of Theo. and Res. for the Soc. of Ed. Greenwood Press.
Cantle, E. (2001). Community Cohesion: A Report of the
Independent Review Team.
Castles, S. & Miller, M. J. (2003). The age of migration, 3rd Ed. Basingstoke, Palgrave Macmillan.
Coleman, J. S. (1990). Foundations of Social Theory.
Commission on Integration and Cohesion (2007). Our shared
future.
Costa, D. L., & Kahn, M. E. (2003). Quarterly Journal of Economics, 118(2), 519-548.
Deffuant, G., et al. (2001), Adv. in Complex Sys., 3:87-98.
Denham, J. (2001). Building Cohesive Communities.
Dorogovtsev, S. N.; Mendes, J.F.F. (2003), "Evolution of Networks",
Ford, R (2008). British Journal of Sociology 59. 609-636.
Gordon, M. (1964). Assimilation and American life.
Halpern, D. (2005). Social Capital.
Li, Y., & Heath, A. (2008). Int. J. Sociology and Social Pol., 28(5-6), 231-244.
Newman, M. et al (2006), "The Structure and Dynamics of Networks", Princeton University Press.
Park, R. E., & Burgess, E. W. (1921/1969). Introduction
to the science of sociology.
Portes, A. (1998). Annual Review of Sociology, 24: 1-24.
Putnam, R (2000) ‘Bowling Alone: The Collapse and Revival of American Community’ Simon and Schuster.
Putnam, R. (2007). Scandinavian Political Studies, 30(2), 137-174.
Riolo, R. L., et al. (2001), Nature, 411:441-443.
Schwarz, N. & Ernst, A. (2009), Technological Forecasting and Social Change, 76(4):497-511.
Skocpol T. (1996). Unraveling from above. American Prospect, 25:20-25.
Strogatz, S. H. (1994) "Nonlinear Dynamics and
Chaos", Perseus Publishing,
van Kampen, N. G. (2007) "Stochastic Processes in Physics and Chemistry", Elsevier.