Frontiers in Sociology and Social Research 5 Series Editor: Richard Serpe Thomas Chesney Agent-Based Modelling of Worker Exploitation Slave from the Machine Frontiers in Sociology and Social Research Volume 5 SeriesEditor RichardSerpe,DepartmentofSociology,KentUniversity,Kent,OH,USA Frontiers of Sociology and Social Research is a cutting-edge social science book series focusing on new directions in sociological and broader social science research. These new directions could be novel theoretical paradigms, developing topical areas of research, innovative methodologies, and/or substantive findings that exemplify and anticipate trends in subfields. The series is predicated on the observationthatanyfieldofknowledgeincontemporarytimesisadynamicrapidly changingbodyofperspectivesandunderstandingthatcontinuouslybuildsuponthe foundation of extant scholarship. The series encourages manuscript submissions from both new and established scholars of sociology, anthropology,social policy, andotherallieddisciplines. Moreinformationaboutthisseriesathttp://www.springer.com/series/8690 Thomas Chesney Agent-Based Modelling of Worker Exploitation Slave from the Machine ThomasChesney UniversityofNottingham Nottingham,UK ISSN2523-3424 ISSN2523-3432 (electronic) FrontiersinSociologyandSocialResearch ISBN978-3-030-75133-3 ISBN978-3-030-75134-0 (eBook) https://doi.org/10.1007/978-3-030-75134-0 ©TheEditor(s)(ifapplicable)andTheAuthor(s),underexclusivelicensetoSpringerNatureSwitzerland AG2021 Thisworkissubjecttocopyright.AllrightsaresolelyandexclusivelylicensedbythePublisher,whether thewhole orpart ofthematerial isconcerned, specifically therights oftranslation, reprinting, reuse ofillustrations, recitation, broadcasting, reproductiononmicrofilmsorinanyotherphysicalway,and transmissionorinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilar ordissimilarmethodologynowknownorhereafterdeveloped. Theuseofgeneraldescriptivenames,registerednames,trademarks,servicemarks,etc.inthispublication doesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfromtherelevant protectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthors,andtheeditorsaresafetoassumethattheadviceandinformationinthisbook arebelievedtobetrueandaccurateatthedateofpublication.Neitherthepublishernortheauthorsor theeditorsgiveawarranty,expressedorimplied,withrespecttothematerialcontainedhereinorforany errorsoromissionsthatmayhavebeenmade.Thepublisherremainsneutralwithregardtojurisdictional claimsinpublishedmapsandinstitutionalaffiliations. ThisSpringerimprintispublishedbytheregisteredcompanySpringerNatureSwitzerlandAG. Theregisteredcompanyaddressis:Gewerbestrasse11,6330Cham,Switzerland Computationalsocialscientistshavea saying: Don’t TrustYourComputer. Well, theydon’t... ...butmaybetheyshould. followingEpstein&Axtell, thisbookisdedicatedto the memoryof myfather,and toJoseph Foreword Research on the topic of sustainability in business and organisations is challeng- ing. Empirical research such as case studies can generate insightful description and analysis of the phenomenon, however, usually without providing tools for experimentation,strategicguidanceordecision-supporttomanagers.Mathematical modellingresearchcoulddeliverthisbutoftenstruggleswithsufficientlygrasping the complexity and context of the subject under investigation. Thus modelling is oftendetachedfromtheempiricalconditionstobecomeapowerfultoolforadvanc- ing transformation towards sustainability. Coming the way of empirical supply chainandoperationsmanagementresearch,Ibecameawareofthisdilemmawhich spurred some methodological wanderings, particularly towards decision-support techniquessuchasanalyticalhierarchyprocess(AHP)andmodellingtoolssuchas systemdynamics(SD)andtoalessextentagent-basedmodelling.Iappreciatedthe increasedaccuracyandconcisenesswhichcome,forexample,withtheformulation of a relationship between two variables as mathematical equation compared to merely linguistic descriptions of this relationship based on words and grammar. At the same time, I strived for a strong empirical grounding of those methods (by basing them on primary and secondary data) and for combining them with other research designs such as qualitative case study research. These endeavours have been first attempts to walk on new grounds that are more appropriate to the normative mission of research for sustainability, and they have not incorporated matureresearchprotocolsyet. It was a coincidence that I met Thomas when I started developingmy interest fortheissueofmodernslaveryattheUniversityofNottingham,UnitedKingdom, in 2014. Then an assistant professor of informationsystems, he was an academic colleague. He had just become passionate about the topic of labour exploitation andmodernslaveryaroundthesametime.TogetherwithAlexanderTrautrims,we shared thoughts and inspiration during occasional campus walks. Similar to the issue of sustainable development of business and society in general, instances of labourexploitationandworkplaceabusearedistinctlycharacterisedbycomplexity deriving from structure, inter-human interaction, and context. In addition, the phenomenaareoftenhiddenbyfactorygates,humanstruggleforsurvival,orguilt ix x Foreword andshame,andthusprimarydatacollectionturnsouttobedifficultandsometimes dangerous and/or ethically problematic. In such complication of things, doing research and finding pathways towards decent work requires new theoretical and methodological angles. Thomas convinced me about the potential of agent-based modelling as a tool of computational social science that is particularly apposite for management and sustainability research. This philosophy of agent modelling applied to workplace abuse as laid out in this book contains several hooks that will enable readers who are interested or already engaged in social science and sustainability research to think twice and deep. This philosophypostulates agent- based modelling as an experimental research design that allows for profound analysisandexperimentalunderstandingofthephenomenonoflabourexploitation, and related strategies of remediation. Furthermore, the book gives a couple of powerful examples of application as well as useful practical guidance on how to designanddoresearchwithagentmodels. Thomas and I have not only shared random walks on Nottingham University Business School’s Jubilee Campus but we have also been engaging in a couple of jointstudies in the broaderfield of managementand labourissues, where I got moreintimatelyacquaintedwiththecharacteristicsofagentmodellingasaresearch method. For me, the power of agent-based modelling resides in its suitability as a complement for and a bridge to other methods as it can be connected to empiricalresearch as well as other modellingtypes such as system dynamics;for example, AnyLogic is such a multi-method simulation tool that connects agent- based modelling with system dynamics and discrete-event process modelling. In this way, agent-based modelling allows for methodological integration and, if applied for validation reasons, methodological triangulation. At the same time, it may help refuting, corroborating or refining theory that has been formulated on conceptual reasoning or empirical data, for example regarding its assumptions, hypotheses and boundary conditions. In this sense, agent-based modelling may stepinascomplementarymethodinsituationswhereempiricalresearchisdifficult or dangerous, as is the case regarding various research problems in the field of modernslavery.Finally,agent-basedmodellingmayalsorepresentapowerfultool forresearchthatprovidesprescriptionandguidancetomanagersandotherdecision makers; for this purpose, we designed agent modelling as a tool of decentralised shadowaccountingthatmaybefedbyindividualobservatorydataofcustomersof UKcarwashesandcouldbeusedbypoliceauthorities. Inanutshell,thecurrentbookopensexcitingavenuesforcomputationalsustain- abilityresearchthroughagent-basedmodels.Itwillinspirereaderstoembarkonto their own exploratory journeys of research on labour exploitation, sustainability managementandbeyond. Kassel,Germany StefanGold 2020 Preface Fewpeopleeverreadtheprefacesothisonewillbeshort.SlavefromtheMachine emerged from a series of agent-based models that were created between summer 2017 and spring 2020 to illustrate how minimum wage abuses might arise in an industry. The work was a research programme within the Rights Lab at the UniversityofNottingham.ThetextisinspiredbyWilliamSimsBainbridge’sbook God from the Machine,but owes a debtto Growing Artificial Societies by Joshua EpsteinandRobertAxtell.MythinkingwasheavilyinfluencedbyEricWinsberg, andcomputercodethatWilliamSimsBainbridgekindlysharedwithmewhenIfirst startedworkingwithagentmodels(itwaswritteninDelphiIthink). The book’spurpose is to illustrate the potential for computersimulation in the studyofmodernslaveryandbyextensionsocialscienceingeneral,andtolayout a philosophy of how it can be used. It will also introduce agent programming to anyreaderwhocarestofollowthecodingsections.Thosethatdowillbelearning NetLogo,apopularandfreeagentmodellingenvironment. Part I presents the philosophy of using agent models to create knowledge. It is a personalphilosophy,one that I have come to from using agent modelling for research,andmuchofwhatissaidisopentodebate.PartIIillustratesitsusewitha seriesofmodelsincludedtopresentaspectsof,andissuesaround,modellingworker exploitation.Theintentionisnotprimarilytodocumentresearchresultsbuttoallow forarichdiscussionofusingagentmodellingasamethod.PartIIIthencoverssome practicalaspectsoffollowingthisresearchmethod. xi