An agentbased simulation Laurence LessardPhillips Institute for Social Change University of Manchester Nick Crossley Department of Sociology University of Manchester Bruce Edmonds Centre for Policy Modelling Manchester Metropolitan University ID: 622503
Download Presentation The PPT/PDF document "Ethnic diversity, density and their cons..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Ethnic diversity, density and their consequences on political participation:An agent-based simulation
Laurence Lessard-Phillips, Institute for Social Change, University of ManchesterNick Crossley, Department of Sociology, University of ManchesterBruce Edmonds, Centre for Policy Modelling, Manchester Metropolitan UniversityEd Fieldhouse, Institute for Social Change, University of ManchesterYaojun Li, Institute for Social Change, University of ManchesterRuth Meyer, Centre for Policy Modelling, Manchester Metropolitan UniversityNick Shryane, Institute for Social Change, University of ManchesterSlide2
Background
Ethnic minorities are (slowly) becoming a bigger part of the UK’s national population~5.6% (1991) – ~7.9% (2001) – ~ 14-18% (2051)Engagement of ethnic minorities in ‘conventional politics’, and its main determinants, is an interesting topic of enquiryEthnic minorities becoming increasingly important segment of the electorateEspecially given their location, density and diversity in the UKLink to socio-political integration/incorporation and other related issues (representation, etc.)UK case peculiar given voting right of Commonwealth citizens
2Slide3
3Slide4
Theorising the role of ethnic diversity and density on turnout
Ethnic diversity (based on Fieldhouse and Cutts, 2008)Group conflict theory: diversity leading to higher levels of conflict and hence mobilisation of the population, leading to higher levels of turnoutCan also have depressing effect on turnoutEconomic resources theory: highly diverse communities have weaker mobilising effects and higher barriers to participation due to lack of resourcesRacial diversity thesis: high levels of diversity display more inequalities and hence lower participationSocial capital theory (?): link between diversity and levels of interpersonal/generalised trust
4Slide5
Theorising the role of ethnic diversity and density on turnout
Ethnic density (based on Fieldhouse and Cutts, 2008)Social capital theory: group concentration leading to higher levels of bonding capital, connectedness and networks, which generate higher levels of political mobilisation and hence turnout Ethnic community model: higher levels of group consciousness/awareness leading to higher levels of turnoutMay also cause alienationRelative deprivation theory: higher levels of deprivation may lead to increased levels of alienation and, in turn, to decreased turnout
5Slide6
Political participation of ethnic minorities in the UK: Existing evidence
Research-based evidence has found divergence in the turnout rates of various ethnic minority groups, with (some) stabilisation of turnout rates over timeAsian turnout > than turnout for non-AsiansDifferentiation within Asian groupsBlack Caribbean and African groups: lower levels amongst ethnic minority groupsYet more recent evidence seems to contradict these claimsSomehow difficult to disentangle ethnic group effects from other effects such as age, socio-economic status, etc.No clear agreement as to the impact of density on ethnic minority turnoutData/methods issues
6Slide7
But…
We are still a long way from understanding this issue without integrating the varied accounts that exist into a unified model that captures the complexity of processes that might be at playLinking the micro to the macroOne way in which you can try to do this is via agent-based simulationUnderutilised method informed by data/evidence/theory in the social sciences that can link multiple and multi-faceted influential processes7Slide8
Agent-based simulation
What is it?Computational description of a given processNot usually analytically tractable More context-dependent…… but assumptions are much less drasticDetail of unfolding processes accessible
more criticisable (including by non-experts)Used to explore inherent possibilities
Validatable by data, opinion, narrative ...Often very complex
What happens?
Entities in simulation are decided up
Behavioural rules for each agent specified
e.g. sets of rules like: if this has happened then do this
Repeatedly evaluated in parallel to see what happens
Outcomes are inspected, graphed, pictured, measured and interpreted in different ways
8Slide9
Dilemmas using this approach
KISS (Keep it Simple, Stupid)Models should be simple enough to understand and check (rigour) May omit critical aspects of the system of interest (lack of relevance)Strong inferences possible about within-model processesWeak mapping to the thing being modelled
KIDS (Keep it Descriptive, Stupid)
Models should capture the critical aspects of social interaction (relevance)
They may be too complex to understand and thoroughly check (lack of rigour)Weak inferences about within-model processesClear mapping to the thing being modelled
9Slide10
Agent-based simulation model of voting behaviour: ‘the’ model
10Slide11
Encapsulating narrative stories of voting in the simulation
Based on collected evidence, we set out stories according to which our agents actE.g.I voted for party X because it will put limits on immigration.I voted for minor party Y because I wanted to send a message to those lying, cheating, fiddlers in Westminster.I always vote – it’s part of who I am.I didn’t vote – what’s the point?Was there an election on?
These narratives also take into account the characteristics of the agents, their dynamics, and other influences
11Slide12
Using the Simulation: the context
Run simulations that correspond to different settings or behavioural hypothesesSee how it affects outcomes, e.g.:Ethnic minority turnoutEthnic majority turnoutResults do not predict, but reveal possible emergent outcomesMore importantlyRaises new questions and gaps in knowledgeAn “in vitro” exploration of some of the complex relationships between factors that can occurSuggests new hypotheses (or refinements on old hypotheses) in an explicit and demonstrated form
12Slide13
Results from preliminary “Proof of concept” version
13Slide14
Early, “Proof of Concept” Version of the Model
Simulation model still being developed Validation stage yet to begin in earnestDemonstrated with 4 different scenariosOnly difference in minorities are (a) those inherent in the data we used to initialise the model and (b) the homophily effect of agents tending to make social links with similar age/ethnicity/politicsModel was run 25 timesAverage turnout in minority and majority is then measured
14Slide15
Turnout by majority, minority (1%IR)Slide16
Turnout by majority, minority (5%IR)Slide17
Conclusion
ABS links micro- and macro-level processes in an explicit manner, enabling the exploration of the effects of how individuals behave and relate to each other at the aggregate levelStill in the process of developing the modelGathering evidenceMaking linkagesUpdating narrativesSpecial focus on extensive exploration of dynamic social networks (e.g., household-level influences)Future development of narrative model “translations”
17Slide18
Thank you!
http://www.scid-project.org18Slide19
What it looks like…
19Slide20
What it looks like…
20Slide21
What it looks like…
21Slide22
What it looks like…
22