When Metrics Govern: Pre-Algorithmic Optimisation and the Ethics of Artificial Intelligence

When Metrics Govern: Pre-Algorithmic Optimisation and the Ethics of Artificial Intelligence | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article When Metrics Govern: Pre-Algorithmic Optimisation and the Ethics of Artificial Intelligence Ifeoluwa E Elemure, Chijioke C. Chuwa, Lukman O. Kolawole, Adeyinka G. Ologun, and 3 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8752951/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Contemporary debates in AI ethics often associate algorithmic governance with advanced computational systems, automated decision-making, and machine learning. This focus, however, risks overlooking how similar governance effects emerge through non-AI optimisation infrastructures that precede and normalise algorithmic control. This paper introduces the concept of algorithmic governance without algorithms to examine how metrics, dashboards, and performance indicators embedded within Lean–Green manufacturing systems reshape authority, agency, and ethical responsibility in organisational life. Drawing on mixed-method empirical material from Nigerian manufacturing organisations—including surveys, semi-structured interviews, and observational data—the study analyses how optimisation regimes function as socio-technical systems of governance rather than neutral managerial tools. The findings reveal three interrelated dynamics: the dominance of quantified metrics in decision-making, the reconfiguration of worker agency through conditional participation, and the transformation of sustainability from a collective ethical commitment into a measurable performance output. These dynamics demonstrate that ethical concerns commonly attributed to artificial intelligence—such as displacement of human judgement, responsibility gaps, and moral deskilling—are already present within everyday organisational technologies. By extending ethical analysis beyond narrowly defined AI systems, this paper contributes to AI ethics scholarship by highlighting the need to address governance risks embedded in pre-algorithmic optimisation infrastructures that shape how institutions value efficiency, responsibility, and human judgement. Algorithmic governance AI ethics Quantification and metrics Optimisation regimes Socio-technical systems Responsibility and agency Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 1. Introduction Artificial intelligence is increasingly understood not only as a set of computational techniques, but as a broader mode of governance that reshapes decision-making, authority, and ethical responsibility across contemporary institutions [ 1 ]. Within debates on AI ethics, growing attention has been paid to algorithmic management, data-driven oversight, and automated optimisation systems that subtly regulate human behaviour while obscuring accountability [ 2 , 3 ]. Yet this focus on advanced or explicitly computational systems risks overlooking a quieter but equally consequential transformation: the spread of optimisation infrastructures—metrics, dashboards, key performance indicators, and performance targets—that produce governance effects typically associated with AI, even in the absence of machine learning or automated decision systems [ 4 , 5 ]. This paper argues that many of the ethical challenges now attributed to artificial intelligence—such as the displacement of judgement, the narrowing of moral responsibility, and the reconfiguration of human agency—are already embedded in everyday organisational technologies that precede, enable, and normalise algorithmic governance [ 5 , 7 ]. However, this dominant framing obscures a more profound transformation taking place within contemporary organisations. Lean–Green systems do not merely improve processes; they increasingly function as infrastructures of governance [ 7 – 9 ]. Through dashboards, key performance indicators (KPIs), audit mechanisms, continuous monitoring, and performance targets, they shape how decisions are made, how authority is exercised, how responsibility is allocated, and whose knowledge counts as legitimate [ 10 , 11 ]. Optimisation, in this sense, is no longer only a technical practice but a mode of organisational regulation [ 12 , 13 ]. While these dynamics are often associated with advanced artificial intelligence and algorithmic management, similar mechanisms operate within organisations long before sophisticated AI is introduced [ 14 ]. Even relatively simple technologies of quantification—such as spreadsheets, dashboards, performance reports, and benchmarking tools—can exert powerful governing effects. When metrics become embedded in everyday routines, they do more than describe reality: they begin to define it [ 15 , 16 ]. They privilege what can be measured, marginalise what cannot, and reshape professional judgement around numerical representations of performance. Critical scholarship on quantification and algorithmic governance has shown that data-driven systems are never neutral. Metrics encode values, reflect political choices, and redistribute power within institutions [ 17 , 18 ]. However, much of this literature focuses on digital platforms, automated systems, and computational algorithms. Far less attention has been paid to how managerial optimisation frameworks—such as Lean–Green systems—already perform analogous governance functions within everyday organisational life, particularly in contexts outside the Global North [ 19 – 21 ]. This gap is especially significant in developing economies, where Lean–Green frameworks are often introduced through modernisation agendas, external consultancy models, and international sustainability standards [ 22 , 23 ]. In such contexts, optimisation systems do not enter organisational vacuums; they interact with existing hierarchies, power relations, and cultural dynamics[ 24 , 25 ]. Understanding how these systems reshape authority, participation, and ethical meaning is therefore not only an academic concern but also a matter of social and organisational consequence [ 26 , 27 ]. This study addresses that gap by examining Lean–Green implementation in Nigerian manufacturing organisations through a critical socio-technical lens. Rather than asking whether these systems “work” in a narrow performance sense, the paper asks: How do Lean–Green optimisation systems reconfigure power and authority within organisations? How do they shape workers’ agency and participation in everyday practice? How do they transform sustainability from a collective ethical commitment into a quantified managerial output? To address these questions, the paper introduces the concept of algorithmic governance without algorithms. This concept captures situations in which governance effects typically associated with algorithmic systems—such as behavioural shaping [ 28 – 30 ], displacement of judgement, and concentration of authority in technical artefacts—emerge through metrics, dashboards, and optimisation infrastructures, even in the absence of advanced AI. It highlights how governance is increasingly embedded in organisational technologies that appear mundane and neutral but exert profound influence over decision-making, accountability, and organisational subjectivity [ 31 , 32 ]. Empirically, the paper draws on mixed-method material from Nigerian manufacturing firms, including survey responses, interviews, and observational data. While the original dataset was collected to explore operational and sustainability outcomes, this study reinterprets the material through a critical lens to illuminate the political, ethical, and cultural dimensions of optimisation practices [ 33 , 34 ]. The focus is therefore not on causal measurement, but on understanding how optimisation systems operate as lived organisational infrastructures. By shifting attention from performance outcomes to governance effects, this paper contributes to growing debates within technology and society scholarship that call for a deeper examination of how systems of optimisation structure human experience, institutional power, and ethical responsibility [ 35 , 36 ]. It argues that Lean–Green frameworks must be understood not only as technical instruments but as socio-technical systems that shape organisational life in ways that deserve critical scrutiny. To capture these dynamics, this paper introduces the concept of algorithmic governance without algorithms, a framework for analysing how optimisation systems generate AI-like governance effects through quantification, performance monitoring, and managerial control rather than through advanced computational intelligence. Drawing on mixed-method empirical material from Lean–Green manufacturing organisations in Nigeria, the study examines how metrics and optimisation regimes restructure authority, condition participation, and transform sustainability from an ethical commitment into a quantified performance outcome. By extending ethical analysis beyond narrowly defined AI systems, the paper contributes to debates in AI ethics by demonstrating that many governance risks associated with artificial intelligence are already present within existing organisational infrastructures. Recognising these pre-algorithmic forms of governance is essential for responsible AI development, as it reveals that ethical harms often emerge not at the moment of automation, but through the gradual normalisation of optimisation logics that shape how institutions value efficiency, responsibility, and human judgement. 2. Theoretical Framing Governance within contemporary organisations is increasingly shaped by technical systems rather than solely by human deliberation. Much of the existing literature describes this shift through the concept of algorithmic governance , where computational systems influence decisions, structure behaviour, and guide resource allocation. However, an exclusive focus on advanced artificial intelligence risks overlooking a longer, quieter transformation: the gradual spread of optimisation logics into organisational spaces once governed primarily by judgement, dialogue, and ethical reflection. Lean–Green systems offer a clear illustration of this transformation. Lean practices introduce extensive measurement regimes, including takt time, throughput, defect rates, cycle efficiency, and process bottlenecks. Green practices extend this measurement culture through environmental indicators such as emissions levels, energy consumption, waste ratios, and resource efficiency. On the surface, these indicators appear to be practical tools for improvement. However, once embedded in dashboards, audit frameworks, performance evaluations, and managerial reporting routines, they begin to shape behaviour in more profound ways. Metrics do not merely describe reality; they begin to define what is valued, what is prioritised, and what is ignored. Three interrelated shifts help explain this transformation. First, there is a shift in epistemic authority. Experiential knowledge—workers' understanding of daily operations, contextual risks, and process realities—often takes a back seat to what numerical indicators show. Managers, too, increasingly rely on dashboards and reports as authoritative representations of reality, even when those representations simplify or distort complex situations. Second, responsibility becomes subtly displaced. When targets are not met, accountability is often directed toward individuals or teams, while the design of the metric system itself remains unquestioned. Structural constraints embedded within the optimisation framework—such as unrealistic targets, resource limitations, or conflicting indicators—tend to be treated as neutral rather than political choices. Third, ethical concerns are translated into operational variables. Concepts such as sustainability, wellbeing, fairness, and participation are no longer primarily treated as moral commitments but are instead reframed as performance categories to be managed. Issues that resist quantification—such as dignity, voice, or long-term social impact—are therefore pushed to the margins. This perspective echoes insights from critical technology studies, which emphasise that technical systems always carry values and assumptions. Lean–Green frameworks, particularly when implemented through rigid metric regimes, often privilege control over care, efficiency over dignity, and output over human flourishing. This does not suggest that such systems are inherently harmful, but rather that they are never neutral: their consequences depend on how they are designed, interpreted, and governed. 3. Methodology This study adopts a qualitative, reinterpreted approach grounded in mixed-method empirical material originally collected from manufacturing organisations in Nigeria. The primary data were generated through a combination of structured questionnaires, semi-structured interviews, and industrial field observations using value stream mapping. While the initial investigation focused on operational performance and sustainability outcomes, this article revisits the same dataset through a socio-ethical perspective, enabling a deeper examination of power, agency, and governance within organisational practice. 3.1. Data Sources Three main sources of evidence inform the analysis. First, survey data were gathered from both shop-floor workers and top-level managers across several manufacturing firms. These instruments explored perceptions of Lean–Green implementation, levels of management commitment, and experiences of participation within organisational improvement initiatives. Second, semi-structured interviews were conducted with managerial personnel, providing insight into implementation challenges, decision-making practices, and organisational culture. These interviews offered valuable contextual depth beyond the structured survey responses. Third, observational material was generated through process-mapping activities and case-study engagement in an automotive manufacturing setting, where value stream mapping documented workflows, interactions, and improvement practices. 3.2. Analytical Strategy Rather than applying statistical or predictive modelling, the study employs an interpretive thematic approach. Survey responses are used descriptively to highlight relational patterns, such as contrasts between managerial narratives and shop-floor experiences. Interview transcripts and observational notes were then analysed to identify recurring themes related to authority, participation, accountability, and control. The analysis is structured around three interrelated dimensions: governance by metrics, which examines how dashboards and performance indicators influence decisions; reconfiguration of agency, which considers how participation is enabled or constrained within optimisation systems; and quantification of ethics, which explores how concepts such as sustainability and fairness become reframed as performance outputs. This methodological approach is particularly suited to AI & Society because it foregrounds meaning, power, and cultural dynamics rather than treating organisational systems as neutral technical mechanisms. 4. Findings The analysis revealed three closely connected patterns across the participating organisations: the growing dominance of metrics in decision-making, the restructuring of worker agency under optimisation regimes, and the transformation of sustainability from a moral commitment into a quantified managerial output. Together, these findings illustrate how Lean–Green systems function not only as technical tools but as socio-technical mechanisms that reshape authority, participation, and ethical meaning within organisational life. 4.1. Governance by Metrics Across all organisations, performance indicators occupied a central position in managerial practice. Managers routinely described organisational success in terms of quantifiable outcomes, such as productivity ratios, defect-reduction rates, efficiency scores, delivery times, and waste-reduction percentages. Management meetings, performance reviews, and operational discussions were often structured around dashboards and performance reports, with numerical indicators serving as the primary reference point for evaluating progress. Several managers acknowledged that their discretion had become increasingly constrained by the presence of these metrics. Decisions that previously relied on contextual judgement—such as adjusting workloads due to worker fatigue, accommodating temporary disruptions in supply chains, or responding flexibly to equipment breakdowns—were now frequently justified or rejected based on whether they aligned with reported indicators. Managers explained that proposals perceived as "not supported by the numbers" were unlikely to be approved, even when they appeared sensible in practice. This reliance on metrics produced a notable paradox. Formally, managers remained responsible for decision-making and performance outcomes. In practice, however, they increasingly described themselves as accountable to the indicators system itself rather than to their own professional judgement. Some managers expressed frustration that the metrics had become an external authority shaping what they could justify, prioritise, or defend. Rather than acting purely as decision-makers, managers were positioned as intermediaries between numerical systems and the workforce. This dynamic reflects a subtle but significant shift in governance. Authority no longer flowed solely from hierarchical position or experience; instead, it was increasingly mediated through quantified representations of performance. The metrics did not merely report organisational activity—they actively shaped managerial reasoning, constrained discretionary judgement, and structured what could be considered legitimate action. 4.2. Reconfigured Worker Agency A second key finding concerned the divergence between managerial narratives of participation and workers lived experiences. Survey responses from managers consistently indicated that Lean–Green initiatives encouraged empowerment, engagement, and inclusion in decision-making. Managers often described improvement programmes as participatory, highlighting mechanisms such as suggestion schemes, team meetings, and continuous improvement activities. However, workers' responses painted a more complex and ambivalent picture. While some workers acknowledged opportunities to contribute ideas, many described participation as conditional rather than genuine. Contributions were encouraged primarily when they aligned with existing productivity targets or efficiency objectives. Workers reported that suggestions to reduce cycle times, improve output, or minimise waste were welcomed, whereas concerns about workload pressure, fatigue, safety risks, or wellbeing were often treated as secondary or dismissed altogether. Several workers indicated that they had learned to frame their contributions in the language of optimisation in order to be heard. Rather than raising issues in terms of fairness or wellbeing, they adjusted their language to emphasise productivity, efficiency, or performance improvement. This adaptation allowed them to participate within the accepted framework, but it also constrained the kinds of concerns they could express. As a result, participation became increasingly performative. Workers were involved in improvement processes, but within narrow boundaries defined by managerial and metric-based priorities. Their agency was not eliminated, but it was reshaped to fit the logic of optimisation. Genuine influence over the design of the system itself remained limited, while responsibility for meeting targets was often individualised at the shop-floor level. This pattern illustrates how optimisation systems do not simply organise work; they actively structure the conditions under which voice, participation, and agency are recognised as legitimate. 4.3. Sustainability as Quantified Output The third significant finding concerns how sustainability was understood and enacted within participating organisations. Sustainability featured prominently in organisational discourse, but almost exclusively through quantitative indicators. Managers frequently referred to reductions in waste percentages, improvements in energy efficiency, compliance scores, audit results, and environmental reporting metrics when describing sustainability achievements. While these measures demonstrated technical progress, they also revealed a narrowing of the conceptualisation of sustainability. Rarely was sustainability discussed alongside worker dignity, organisational justice, community impact, or long-term social responsibility. Discussions about sustainability were primarily framed around what could be measured, reported, and displayed on dashboards. This reflects a deeper transformation in meaning. Sustainability was no longer treated primarily as an ethical or collective project involving negotiation about values, responsibilities, and long-term consequences. Instead, it became a performance outcome to be demonstrated through numerical indicators. Practices were judged not by whether they were experienced as fair or humane, but by whether they produced favourable metrics. In this way, sustainability was effectively operationalised. What could be quantified gained legitimacy, while broader ethical concerns that resisted measurement became marginal. This shift illustrates how optimisation systems shape not only organisational practices but also the very meanings attached to concepts such as responsibility, care, and sustainability. 5. Discussion The findings of this study indicate that Lean–Green frameworks operate not only as operational improvement tools but also as infrastructures of governance within manufacturing organisations. Their influence extends beyond productivity and environmental outcomes to shape organisational subjectivities—how managers understand their roles, how workers experience participation, and how sustainability itself is conceptualised and enacted. In this sense, Lean–Green systems function as socio-technical arrangements that structure power, responsibility, and meaning across organisational life. This perspective challenges the widespread assumption that efficiency-oriented frameworks are inherently neutral. Performance indicators, dashboards, and targets are often presented as objective representations of reality. However, the findings demonstrate that every metric reflects implicit value judgements about what matters and what does not. When organisations privilege numerical indicators, they inevitably prioritise specific values—such as speed, output, and compliance—while marginalising others, including care, deliberation, dignity, and justice. What appears as technical rationality is therefore deeply normative, shaping behaviour and legitimising particular forms of organisational conduct. The study also complicates dominant narratives of empowerment frequently associated with Lean implementation. Participation is commonly framed as a central feature of continuous improvement cultures, yet the evidence suggests that participation is often tightly structured by optimisation logics. Workers are invited to contribute ideas, but primarily within predefined boundaries that align with productivity goals and performance targets. Contributions that question workload expectations, safety concerns, or the fairness of targets are less likely to gain traction. As a result, agency becomes conditional rather than genuine. Meaningful empowerment would require creating spaces where workers can challenge the system's design, not merely optimise their performance within it. A further implication concerns the evolving meaning of sustainability within organisations. The findings reveal that sustainability is increasingly understood in terms of measurable outputs rather than ethical deliberation. Waste reduction percentages, energy efficiency indicators, and compliance scores become proxies for responsibility, while broader concerns about wellbeing, equity, and long-term societal impact receive limited attention. These risks hollow out the ethical foundations of sustainability, reducing it to a reporting exercise rather than a collective moral commitment. When sustainability becomes something to be displayed on dashboards rather than negotiated among stakeholders, its transformative potential is weakened. Importantly, these dynamics resonate strongly with contemporary critiques of algorithmic governance. Scholars have raised concerns that data-driven systems can subtly shape decision-making, constrain agency, and redistribute responsibility in ways that are difficult to contest. The present study demonstrates that similar processes are already embedded within managerial technologies such as metrics, dashboards, and audit systems. Governance, in this context, is not imposed by advanced artificial intelligence but emerges from everyday organisational infrastructures that encode optimisation logics into practice. By revealing how optimisation systems already enact forms of algorithmic governance through metrics and performance regimes, this study underscores that the ethical risks often attributed to artificial intelligence—such as responsibility displacement, constrained agency, and moral deskilling—are not future concerns but present realities embedded in contemporary organisational infrastructures, demanding ethical scrutiny well before the introduction of advanced AI. 6. Conclusion This study demonstrates that Lean–Green optimisation systems cannot be understood as neutral managerial tools but must be recognised as socio-technical infrastructures of governance that reshape authority, agency, and ethical responsibility within organisations. By introducing the concept of algorithmic governance without algorithms, the paper shows how metrics, dashboards, and performance indicators produce governance effects commonly associated with artificial intelligence, even in the absence of advanced computational systems. Through empirical evidence from Nigerian manufacturing organisations, the analysis reveals how quantified optimisation regimes displace human judgement, condition participation, and transform sustainability from a collective ethical commitment into a measurable performance outcome. These findings contribute to AI ethics by extending ethical scrutiny beyond narrowly defined AI technologies to the broader organisational infrastructures that normalise algorithmic modes of control. They highlight that many ethical risks attributed to future AI systems—such as responsibility gaps, moral deskilling, and constrained human agency—are already embedded in everyday optimisation practices. Recognising these pre-algorithmic forms of governance is essential for responsible AI development, as it underscores that ethical AI cannot be achieved solely through technical safeguards but requires critical engagement with the values, assumptions, and power relations encoded within the systems that shape organisational decision-making long before automation occurs. Declarations Conflicts of interest The authors declare no conflict of interest. Ethics approval Ethical approval was obtained as part of the original study, and all participants provided informed consent. Consent to participate Informed consent was obtained from all participants involved in the study. Funding The authors received no specific funding for this study. Author Contribution I.E.E. conceived the study, developed the theoretical framework, and led the writing of the manuscript. C.C.C. and L.O.K. contributed to the conceptual development, literature review, and critical interpretation of findings. A.G.O. and A.F.I. supported the research design, data interpretation, and methodological refinement. G.A.E. contributed to empirical analysis and contextual interpretation of organisational data. O.M.I. contributed to the ethical framing, governance analysis, and critical revisions of the manuscript. All authors reviewed, edited, and approved the final version of the manuscript and agree to be accountable for all aspects of the work. Acknowledgement none References Danaher, J.: Algorithmic governance: Developing a research agenda. Big Data Soc. 4 (2), 1–21 (2017). https://doi.org/10.1177/2053951717726554 Ziewitz, M.: Governing algorithms: Myth, mess, and methods. Sci. Technol. Hum. Values. 41 (1), 3–16 (2016). https://doi.org/10.1177/0162243915608947 Pasquale, F.: The black box society: The secret algorithms that control money and information. Harvard University Press (2015) Kitchin, R.: Thinking critically about and researching algorithms. Inform. Communication Soc. 20 (1), 14–29 (2017) Gillespie, T.: The relevance of algorithms. In: Gillespie, T., et al. (eds.) Media technologies, pp. 167–194. MIT Press (2014) Beer, D.: The social power of algorithms. Inform. Communication Soc. 20 (1), 1–13 (2017) Espeland, W.N., Stevens, M.L.: Commensuration as a social process. Ann. Rev. Sociol. 24 , 313–343 (1998) Espeland, W.N., Stevens, M.L.: A sociology of quantification. Eur. J. Sociol. 49 (3), 401–436 (2008) Power, M.: The audit society: Rituals of verification, 2nd edn. Oxford University Press (2021) Desrosières, A.: The politics of large numbers: A history of statistical reasoning. Harvard University Press (1998) Porter, T.M.: Trust in numbers: The pursuit of objectivity in science and public life. Princeton University Press (1995) O’Neil, C.: Weapons of math destruction. Crown Publishing (2016) Zuboff, S.: The age of surveillance capitalism. Profile Books (2019) Amoore, L.: Cloud ethics: Algorithms and the attributes of ourselves and others. Duke University Press (2020) Eubanks, V.: Automating inequality. St. Martin’s (2018) Kellogg, K.C.: Subtle shifts in control: Professional autonomy and algorithmic oversight. Adm. Sci. Q. 66 (1), 1–41 (2021) Jarrahi, M.H.: Artificial intelligence and the future of work. Bus. Horiz. 61 (4), 577–586 (2018) Orlikowski, W.J.: Sociomaterial practices. Organ. Stud. 28 (9), 1435–1448 (2007) Leonardi, P.M.: When flexible routines meet flexible technologies. MIS Q. 35 (1), 147–167 (2011) Suchman, L.: Human–machine reconfigurations. Cambridge University Press (2007) Winner, L.: Do artifacts have politics? Daedalus. 109 (1), 121–136 (1980) Feenberg, A.: Transforming technology, 2nd edn. Oxford University Press (2010) Flyvbjerg, B.: Making social science matter. Cambridge University Press (2001) Scott, J.C.: Seeing like a state. Yale University Press (1998) Burrell, G., Morgan, G.: Sociological paradigms and organisational analysis. Heinemann (1979) Townley, B.: Performance appraisal and the emergence of management. Acc. Organ. Soc. 18 (3), 221–238 (1993) Roberts, J.: The possibilities of accountability. Acc. Organ. Soc. 16 (4), 355–368 (1991) Grey, C.: Career as a project of the self. Sociology. 28 (2), 479–497 (1994) Alvesson, M., Kärreman, D.: Varieties of discourse. Hum. Relat. 53 (9), 1125–1149 (2000) Willmott, H.: Strength is ignorance; slavery is freedom. J. Manage. Stud. 30 (4), 515–552 (1993) Boltanski, L., Chiapello, È.: The new spirit of capitalism. Verso (2005) Dourish, P.: Algorithms and their others. Big Data Soc., 3 (2). (2016) Latour, B.: Reassembling the social. Oxford University Press (2005) Callon, M.: The laws of the markets. Blackwell, Oxford (1998) Bowker, G.C., Star, S.L.: Sorting things out: Classification and its consequences. MIT Press (1999) Selznick, P.: Leadership in administration. Harper & Row (1957) Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8752951","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":584408547,"identity":"15a29a19-7305-4c24-9e9e-18b4831f021b","order_by":0,"name":"Ifeoluwa E Elemure","email":"","orcid":"","institution":"University of Portsmouth","correspondingAuthor":false,"prefix":"","firstName":"Ifeoluwa","middleName":"E","lastName":"Elemure","suffix":""},{"id":584408548,"identity":"4283f0df-9c41-4586-b47d-669e6ca1ef19","order_by":1,"name":"Chijioke C. Chuwa","email":"","orcid":"","institution":"Northumbria University","correspondingAuthor":false,"prefix":"","firstName":"Chijioke","middleName":"C.","lastName":"Chuwa","suffix":""},{"id":584408550,"identity":"451517d4-63eb-4d72-8cb7-f779e492577f","order_by":2,"name":"Lukman O. Kolawole","email":"","orcid":"","institution":"Middlesex University","correspondingAuthor":false,"prefix":"","firstName":"Lukman","middleName":"O.","lastName":"Kolawole","suffix":""},{"id":584408554,"identity":"954ab9d6-ea4a-4bfb-9b53-d2f9cdcc00bb","order_by":3,"name":"Adeyinka G. Ologun","email":"data:image/png;base64,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","orcid":"","institution":"University of Wolverhampton","correspondingAuthor":true,"prefix":"","firstName":"Adeyinka","middleName":"G.","lastName":"Ologun","suffix":""},{"id":584408555,"identity":"dae1e3ea-21a3-41e8-bdab-a97c0d70bd3e","order_by":4,"name":"Abiodun F Ibidunmoye","email":"","orcid":"","institution":"University of Lagos","correspondingAuthor":false,"prefix":"","firstName":"Abiodun","middleName":"F","lastName":"Ibidunmoye","suffix":""},{"id":584408559,"identity":"0c83a505-62e3-4ca2-824c-4ef7eac0667d","order_by":5,"name":"Grace A. Eneano","email":"","orcid":"","institution":"University of South Wales","correspondingAuthor":false,"prefix":"","firstName":"Grace","middleName":"A.","lastName":"Eneano","suffix":""},{"id":584408561,"identity":"acd4feed-3976-4625-bf54-e43f5d9958a2","order_by":6,"name":"Ogechi M. Ikeakaonwu","email":"","orcid":"","institution":"University of Dundee","correspondingAuthor":false,"prefix":"","firstName":"Ogechi","middleName":"M.","lastName":"Ikeakaonwu","suffix":""}],"badges":[],"createdAt":"2026-02-01 00:23:29","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8752951/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8752951/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":102332079,"identity":"391d2909-6c96-4149-829d-5bd3b6b5fe7a","added_by":"auto","created_at":"2026-02-10 15:26:35","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":731189,"visible":true,"origin":"","legend":"\u003cp\u003eConceptual framework: from lean–green systems to algorithmic governance\u003c/p\u003e","description":"","filename":"floatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8752951/v1/9b5c8f75e888cec3f78f660c.png"},{"id":102332146,"identity":"0ed428bf-90d3-4e76-bb33-8fc9c6847270","added_by":"auto","created_at":"2026-02-10 15:26:50","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1282273,"visible":true,"origin":"","legend":"\u003cp\u003ePower Shift Map: How Authority Moves from Humans to Metrics\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8752951/v1/1721aa533d146f86376897ff.png"},{"id":102332124,"identity":"4e505b05-2ffd-47da-b07e-f45668a62f70","added_by":"auto","created_at":"2026-02-10 15:26:46","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":882037,"visible":true,"origin":"","legend":"\u003cp\u003eDivergence in Perception: Management vs Shop-Floor Experience\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8752951/v1/e3c98c35d8da7e9d7c85c8c4.png"},{"id":102332077,"identity":"14e46123-bb23-4608-be54-9ec64a5712f4","added_by":"auto","created_at":"2026-02-10 15:26:33","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1520797,"visible":true,"origin":"","legend":"\u003cp\u003eThe Transformation of Sustainability: From Ethical Ideal to Quantified Output\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8752951/v1/f8edcf4741a171a6ee0587d6.png"},{"id":102397774,"identity":"99ec49ce-8595-4744-be15-d9790ca1651e","added_by":"auto","created_at":"2026-02-11 10:19:42","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":1807411,"visible":true,"origin":"","legend":"\u003cp\u003eOrganisational Subjectivity Model: How Optimisation Shapes Behaviour\u003c/p\u003e","description":"","filename":"floatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-8752951/v1/8402acd0a1f199de6e901b26.png"},{"id":104901319,"identity":"2c956ef1-1081-44b8-9009-2e7f9a0dba8c","added_by":"auto","created_at":"2026-03-18 13:12:38","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":6327378,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8752951/v1/e266eedb-2b97-4796-b3d1-39da44e0bece.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"When Metrics Govern: Pre-Algorithmic Optimisation and the Ethics of Artificial Intelligence","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eArtificial intelligence is increasingly understood not only as a set of computational techniques, but as a broader mode of governance that reshapes decision-making, authority, and ethical responsibility across contemporary institutions [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Within debates on AI ethics, growing attention has been paid to algorithmic management, data-driven oversight, and automated optimisation systems that subtly regulate human behaviour while obscuring accountability [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Yet this focus on advanced or explicitly computational systems risks overlooking a quieter but equally consequential transformation: the spread of optimisation infrastructures\u0026mdash;metrics, dashboards, key performance indicators, and performance targets\u0026mdash;that produce governance effects typically associated with AI, even in the absence of machine learning or automated decision systems [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. This paper argues that many of the ethical challenges now attributed to artificial intelligence\u0026mdash;such as the displacement of judgement, the narrowing of moral responsibility, and the reconfiguration of human agency\u0026mdash;are already embedded in everyday organisational technologies that precede, enable, and normalise algorithmic governance [\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHowever, this dominant framing obscures a more profound transformation taking place within contemporary organisations. Lean\u0026ndash;Green systems do not merely improve processes; they increasingly function as infrastructures of governance [\u003cspan additionalcitationids=\"CR8\" citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Through dashboards, key performance indicators (KPIs), audit mechanisms, continuous monitoring, and performance targets, they shape how decisions are made, how authority is exercised, how responsibility is allocated, and whose knowledge counts as legitimate [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e, \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]. Optimisation, in this sense, is no longer only a technical practice but a mode of organisational regulation [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eWhile these dynamics are often associated with advanced artificial intelligence and algorithmic management, similar mechanisms operate within organisations long before sophisticated AI is introduced [\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. Even relatively simple technologies of quantification\u0026mdash;such as spreadsheets, dashboards, performance reports, and benchmarking tools\u0026mdash;can exert powerful governing effects. When metrics become embedded in everyday routines, they do more than describe reality: they begin to define it [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. They privilege what can be measured, marginalise what cannot, and reshape professional judgement around numerical representations of performance.\u003c/p\u003e \u003cp\u003eCritical scholarship on quantification and algorithmic governance has shown that data-driven systems are never neutral. Metrics encode values, reflect political choices, and redistribute power within institutions [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. However, much of this literature focuses on digital platforms, automated systems, and computational algorithms. Far less attention has been paid to how managerial optimisation frameworks\u0026mdash;such as Lean\u0026ndash;Green systems\u0026mdash;already perform analogous governance functions within everyday organisational life, particularly in contexts outside the Global North [\u003cspan additionalcitationids=\"CR20\" citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThis gap is especially significant in developing economies, where Lean\u0026ndash;Green frameworks are often introduced through modernisation agendas, external consultancy models, and international sustainability standards [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. In such contexts, optimisation systems do not enter organisational vacuums; they interact with existing hierarchies, power relations, and cultural dynamics[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Understanding how these systems reshape authority, participation, and ethical meaning is therefore not only an academic concern but also a matter of social and organisational consequence [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThis study addresses that gap by examining Lean\u0026ndash;Green implementation in Nigerian manufacturing organisations through a critical socio-technical lens. Rather than asking whether these systems \u0026ldquo;work\u0026rdquo; in a narrow performance sense, the paper asks: How do Lean\u0026ndash;Green optimisation systems reconfigure power and authority within organisations? How do they shape workers\u0026rsquo; agency and participation in everyday practice? How do they transform sustainability from a collective ethical commitment into a quantified managerial output?\u003c/p\u003e \u003cp\u003eTo address these questions, the paper introduces the concept of algorithmic governance without algorithms. This concept captures situations in which governance effects typically associated with algorithmic systems\u0026mdash;such as behavioural shaping [\u003cspan additionalcitationids=\"CR29\" citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e], displacement of judgement, and concentration of authority in technical artefacts\u0026mdash;emerge through metrics, dashboards, and optimisation infrastructures, even in the absence of advanced AI. It highlights how governance is increasingly embedded in organisational technologies that appear mundane and neutral but exert profound influence over decision-making, accountability, and organisational subjectivity [\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eEmpirically, the paper draws on mixed-method material from Nigerian manufacturing firms, including survey responses, interviews, and observational data. While the original dataset was collected to explore operational and sustainability outcomes, this study reinterprets the material through a critical lens to illuminate the political, ethical, and cultural dimensions of optimisation practices [\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. The focus is therefore not on causal measurement, but on understanding how optimisation systems operate as lived organisational infrastructures.\u003c/p\u003e \u003cp\u003eBy shifting attention from performance outcomes to governance effects, this paper contributes to growing debates within technology and society scholarship that call for a deeper examination of how systems of optimisation structure human experience, institutional power, and ethical responsibility [\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. It argues that Lean\u0026ndash;Green frameworks must be understood not only as technical instruments but as socio-technical systems that shape organisational life in ways that deserve critical scrutiny.\u003c/p\u003e \u003cp\u003eTo capture these dynamics, this paper introduces the concept of algorithmic governance without algorithms, a framework for analysing how optimisation systems generate AI-like governance effects through quantification, performance monitoring, and managerial control rather than through advanced computational intelligence. Drawing on mixed-method empirical material from Lean\u0026ndash;Green manufacturing organisations in Nigeria, the study examines how metrics and optimisation regimes restructure authority, condition participation, and transform sustainability from an ethical commitment into a quantified performance outcome. By extending ethical analysis beyond narrowly defined AI systems, the paper contributes to debates in AI ethics by demonstrating that many governance risks associated with artificial intelligence are already present within existing organisational infrastructures. Recognising these pre-algorithmic forms of governance is essential for responsible AI development, as it reveals that ethical harms often emerge not at the moment of automation, but through the gradual normalisation of optimisation logics that shape how institutions value efficiency, responsibility, and human judgement.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"2. Theoretical Framing","content":"\u003cp\u003eGovernance within contemporary organisations is increasingly shaped by technical systems rather than solely by human deliberation. Much of the existing literature describes this shift through the concept of \u003cem\u003ealgorithmic governance\u003c/em\u003e, where computational systems influence decisions, structure behaviour, and guide resource allocation. However, an exclusive focus on advanced artificial intelligence risks overlooking a longer, quieter transformation: the gradual spread of optimisation logics into organisational spaces once governed primarily by judgement, dialogue, and ethical reflection.\u003c/p\u003e \u003cp\u003eLean\u0026ndash;Green systems offer a clear illustration of this transformation. Lean practices introduce extensive measurement regimes, including takt time, throughput, defect rates, cycle efficiency, and process bottlenecks. Green practices extend this measurement culture through environmental indicators such as emissions levels, energy consumption, waste ratios, and resource efficiency. On the surface, these indicators appear to be practical tools for improvement. However, once embedded in dashboards, audit frameworks, performance evaluations, and managerial reporting routines, they begin to shape behaviour in more profound ways. Metrics do not merely describe reality; they begin to define what is valued, what is prioritised, and what is ignored.\u003c/p\u003e \u003cp\u003eThree interrelated shifts help explain this transformation. First, there is a shift in epistemic authority. Experiential knowledge\u0026mdash;workers' understanding of daily operations, contextual risks, and process realities\u0026mdash;often takes a back seat to what numerical indicators show. Managers, too, increasingly rely on dashboards and reports as authoritative representations of reality, even when those representations simplify or distort complex situations.\u003c/p\u003e \u003cp\u003eSecond, responsibility becomes subtly displaced. When targets are not met, accountability is often directed toward individuals or teams, while the design of the metric system itself remains unquestioned. Structural constraints embedded within the optimisation framework\u0026mdash;such as unrealistic targets, resource limitations, or conflicting indicators\u0026mdash;tend to be treated as neutral rather than political choices.\u003c/p\u003e \u003cp\u003eThird, ethical concerns are translated into operational variables. Concepts such as sustainability, wellbeing, fairness, and participation are no longer primarily treated as moral commitments but are instead reframed as performance categories to be managed. Issues that resist quantification\u0026mdash;such as dignity, voice, or long-term social impact\u0026mdash;are therefore pushed to the margins.\u003c/p\u003e \u003cp\u003eThis perspective echoes insights from critical technology studies, which emphasise that technical systems always carry values and assumptions. Lean\u0026ndash;Green frameworks, particularly when implemented through rigid metric regimes, often privilege control over care, efficiency over dignity, and output over human flourishing. This does not suggest that such systems are inherently harmful, but rather that they are never neutral: their consequences depend on how they are designed, interpreted, and governed.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"3. Methodology","content":"\u003cp\u003eThis study adopts a qualitative, reinterpreted approach grounded in mixed-method empirical material originally collected from manufacturing organisations in Nigeria. The primary data were generated through a combination of structured questionnaires, semi-structured interviews, and industrial field observations using value stream mapping. While the initial investigation focused on operational performance and sustainability outcomes, this article revisits the same dataset through a socio-ethical perspective, enabling a deeper examination of power, agency, and governance within organisational practice.\u003c/p\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003e3.1. Data Sources\u003c/h2\u003e \u003cp\u003eThree main sources of evidence inform the analysis. First, survey data were gathered from both shop-floor workers and top-level managers across several manufacturing firms. These instruments explored perceptions of Lean\u0026ndash;Green implementation, levels of management commitment, and experiences of participation within organisational improvement initiatives. Second, semi-structured interviews were conducted with managerial personnel, providing insight into implementation challenges, decision-making practices, and organisational culture. These interviews offered valuable contextual depth beyond the structured survey responses. Third, observational material was generated through process-mapping activities and case-study engagement in an automotive manufacturing setting, where value stream mapping documented workflows, interactions, and improvement practices.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003e3.2. Analytical Strategy\u003c/h2\u003e \u003cp\u003eRather than applying statistical or predictive modelling, the study employs an interpretive thematic approach. Survey responses are used descriptively to highlight relational patterns, such as contrasts between managerial narratives and shop-floor experiences. Interview transcripts and observational notes were then analysed to identify recurring themes related to authority, participation, accountability, and control.\u003c/p\u003e \u003cp\u003eThe analysis is structured around three interrelated dimensions: governance by metrics, which examines how dashboards and performance indicators influence decisions; reconfiguration of agency, which considers how participation is enabled or constrained within optimisation systems; and quantification of ethics, which explores how concepts such as sustainability and fairness become reframed as performance outputs.\u003c/p\u003e \u003cp\u003eThis methodological approach is particularly suited to \u003cem\u003eAI \u0026amp; Society\u003c/em\u003e because it foregrounds meaning, power, and cultural dynamics rather than treating organisational systems as neutral technical mechanisms.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"4. Findings","content":"\u003cp\u003eThe analysis revealed three closely connected patterns across the participating organisations: the growing dominance of metrics in decision-making, the restructuring of worker agency under optimisation regimes, and the transformation of sustainability from a moral commitment into a quantified managerial output. Together, these findings illustrate how Lean\u0026ndash;Green systems function not only as technical tools but as socio-technical mechanisms that reshape authority, participation, and ethical meaning within organisational life.\u003c/p\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003e4.1. Governance by Metrics\u003c/h2\u003e \u003cp\u003eAcross all organisations, performance indicators occupied a central position in managerial practice. Managers routinely described organisational success in terms of quantifiable outcomes, such as productivity ratios, defect-reduction rates, efficiency scores, delivery times, and waste-reduction percentages. Management meetings, performance reviews, and operational discussions were often structured around dashboards and performance reports, with numerical indicators serving as the primary reference point for evaluating progress.\u003c/p\u003e \u003cp\u003eSeveral managers acknowledged that their discretion had become increasingly constrained by the presence of these metrics. Decisions that previously relied on contextual judgement\u0026mdash;such as adjusting workloads due to worker fatigue, accommodating temporary disruptions in supply chains, or responding flexibly to equipment breakdowns\u0026mdash;were now frequently justified or rejected based on whether they aligned with reported indicators. Managers explained that proposals perceived as \"not supported by the numbers\" were unlikely to be approved, even when they appeared sensible in practice.\u003c/p\u003e \u003cp\u003eThis reliance on metrics produced a notable paradox. Formally, managers remained responsible for decision-making and performance outcomes. In practice, however, they increasingly described themselves as accountable to the indicators system itself rather than to their own professional judgement. Some managers expressed frustration that the metrics had become an external authority shaping what they could justify, prioritise, or defend. Rather than acting purely as decision-makers, managers were positioned as intermediaries between numerical systems and the workforce.\u003c/p\u003e \u003cp\u003eThis dynamic reflects a subtle but significant shift in governance. Authority no longer flowed solely from hierarchical position or experience; instead, it was increasingly mediated through quantified representations of performance. The metrics did not merely report organisational activity\u0026mdash;they actively shaped managerial reasoning, constrained discretionary judgement, and structured what could be considered legitimate action.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003e4.2. Reconfigured Worker Agency\u003c/h2\u003e \u003cp\u003eA second key finding concerned the divergence between managerial narratives of participation and workers lived experiences. Survey responses from managers consistently indicated that Lean\u0026ndash;Green initiatives encouraged empowerment, engagement, and inclusion in decision-making. Managers often described improvement programmes as participatory, highlighting mechanisms such as suggestion schemes, team meetings, and continuous improvement activities.\u003c/p\u003e \u003cp\u003eHowever, workers' responses painted a more complex and ambivalent picture. While some workers acknowledged opportunities to contribute ideas, many described participation as conditional rather than genuine. Contributions were encouraged primarily when they aligned with existing productivity targets or efficiency objectives. Workers reported that suggestions to reduce cycle times, improve output, or minimise waste were welcomed, whereas concerns about workload pressure, fatigue, safety risks, or wellbeing were often treated as secondary or dismissed altogether.\u003c/p\u003e \u003cp\u003eSeveral workers indicated that they had learned to frame their contributions in the language of optimisation in order to be heard. Rather than raising issues in terms of fairness or wellbeing, they adjusted their language to emphasise productivity, efficiency, or performance improvement. This adaptation allowed them to participate within the accepted framework, but it also constrained the kinds of concerns they could express.\u003c/p\u003e \u003cp\u003eAs a result, participation became increasingly performative. Workers were involved in improvement processes, but within narrow boundaries defined by managerial and metric-based priorities. Their agency was not eliminated, but it was reshaped to fit the logic of optimisation. Genuine influence over the design of the system itself remained limited, while responsibility for meeting targets was often individualised at the shop-floor level.\u003c/p\u003e \u003cp\u003eThis pattern illustrates how optimisation systems do not simply organise work; they actively structure the conditions under which voice, participation, and agency are recognised as legitimate.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec9\" class=\"Section2\"\u003e \u003ch2\u003e4.3. Sustainability as Quantified Output\u003c/h2\u003e \u003cp\u003eThe third significant finding concerns how sustainability was understood and enacted within participating organisations. Sustainability featured prominently in organisational discourse, but almost exclusively through quantitative indicators. Managers frequently referred to reductions in waste percentages, improvements in energy efficiency, compliance scores, audit results, and environmental reporting metrics when describing sustainability achievements.\u003c/p\u003e \u003cp\u003eWhile these measures demonstrated technical progress, they also revealed a narrowing of the conceptualisation of sustainability. Rarely was sustainability discussed alongside worker dignity, organisational justice, community impact, or long-term social responsibility. Discussions about sustainability were primarily framed around what could be measured, reported, and displayed on dashboards.\u003c/p\u003e \u003cp\u003eThis reflects a deeper transformation in meaning. Sustainability was no longer treated primarily as an ethical or collective project involving negotiation about values, responsibilities, and long-term consequences. Instead, it became a performance outcome to be demonstrated through numerical indicators. Practices were judged not by whether they were experienced as fair or humane, but by whether they produced favourable metrics.\u003c/p\u003e \u003cp\u003eIn this way, sustainability was effectively operationalised. What could be quantified gained legitimacy, while broader ethical concerns that resisted measurement became marginal. This shift illustrates how optimisation systems shape not only organisational practices but also the very meanings attached to concepts such as responsibility, care, and sustainability.\u003c/p\u003e \u003c/div\u003e"},{"header":"5. Discussion","content":"\u003cp\u003eThe findings of this study indicate that Lean\u0026ndash;Green frameworks operate not only as operational improvement tools but also as infrastructures of governance within manufacturing organisations. Their influence extends beyond productivity and environmental outcomes to shape organisational subjectivities\u0026mdash;how managers understand their roles, how workers experience participation, and how sustainability itself is conceptualised and enacted. In this sense, Lean\u0026ndash;Green systems function as socio-technical arrangements that structure power, responsibility, and meaning across organisational life.\u003c/p\u003e \u003cp\u003eThis perspective challenges the widespread assumption that efficiency-oriented frameworks are inherently neutral. Performance indicators, dashboards, and targets are often presented as objective representations of reality. However, the findings demonstrate that every metric reflects implicit value judgements about what matters and what does not. When organisations privilege numerical indicators, they inevitably prioritise specific values\u0026mdash;such as speed, output, and compliance\u0026mdash;while marginalising others, including care, deliberation, dignity, and justice. What appears as technical rationality is therefore deeply normative, shaping behaviour and legitimising particular forms of organisational conduct.\u003c/p\u003e \u003cp\u003eThe study also complicates dominant narratives of empowerment frequently associated with Lean implementation. Participation is commonly framed as a central feature of continuous improvement cultures, yet the evidence suggests that participation is often tightly structured by optimisation logics. Workers are invited to contribute ideas, but primarily within predefined boundaries that align with productivity goals and performance targets. Contributions that question workload expectations, safety concerns, or the fairness of targets are less likely to gain traction. As a result, agency becomes conditional rather than genuine. Meaningful empowerment would require creating spaces where workers can challenge the system's design, not merely optimise their performance within it.\u003c/p\u003e \u003cp\u003eA further implication concerns the evolving meaning of sustainability within organisations. The findings reveal that sustainability is increasingly understood in terms of measurable outputs rather than ethical deliberation. Waste reduction percentages, energy efficiency indicators, and compliance scores become proxies for responsibility, while broader concerns about wellbeing, equity, and long-term societal impact receive limited attention. These risks hollow out the ethical foundations of sustainability, reducing it to a reporting exercise rather than a collective moral commitment. When sustainability becomes something to be displayed on dashboards rather than negotiated among stakeholders, its transformative potential is weakened.\u003c/p\u003e \u003cp\u003eImportantly, these dynamics resonate strongly with contemporary critiques of algorithmic governance. Scholars have raised concerns that data-driven systems can subtly shape decision-making, constrain agency, and redistribute responsibility in ways that are difficult to contest. The present study demonstrates that similar processes are already embedded within managerial technologies such as metrics, dashboards, and audit systems. Governance, in this context, is not imposed by advanced artificial intelligence but emerges from everyday organisational infrastructures that encode optimisation logics into practice.\u003c/p\u003e \u003cp\u003eBy revealing how optimisation systems already enact forms of algorithmic governance through metrics and performance regimes, this study underscores that the ethical risks often attributed to artificial intelligence\u0026mdash;such as responsibility displacement, constrained agency, and moral deskilling\u0026mdash;are not future concerns but present realities embedded in contemporary organisational infrastructures, demanding ethical scrutiny well before the introduction of advanced AI.\u003c/p\u003e"},{"header":"6. Conclusion","content":"\u003cp\u003eThis study demonstrates that Lean\u0026ndash;Green optimisation systems cannot be understood as neutral managerial tools but must be recognised as socio-technical infrastructures of governance that reshape authority, agency, and ethical responsibility within organisations. By introducing the concept of algorithmic governance without algorithms, the paper shows how metrics, dashboards, and performance indicators produce governance effects commonly associated with artificial intelligence, even in the absence of advanced computational systems. Through empirical evidence from Nigerian manufacturing organisations, the analysis reveals how quantified optimisation regimes displace human judgement, condition participation, and transform sustainability from a collective ethical commitment into a measurable performance outcome.\u003c/p\u003e \u003cp\u003eThese findings contribute to AI ethics by extending ethical scrutiny beyond narrowly defined AI technologies to the broader organisational infrastructures that normalise algorithmic modes of control. They highlight that many ethical risks attributed to future AI systems\u0026mdash;such as responsibility gaps, moral deskilling, and constrained human agency\u0026mdash;are already embedded in everyday optimisation practices. Recognising these pre-algorithmic forms of governance is essential for responsible AI development, as it underscores that ethical AI cannot be achieved solely through technical safeguards but requires critical engagement with the values, assumptions, and power relations encoded within the systems that shape organisational decision-making long before automation occurs.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflicts of interest\u003c/h2\u003e \u003cp\u003e \u003cem\u003eThe authors declare no conflict of interest.\u003c/em\u003e \u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eEthics approval\u003c/strong\u003e \u003cp\u003e \u003cem\u003eEthical approval was obtained as part of the original study, and all participants provided informed consent.\u003c/em\u003e \u003c/p\u003e \u003c/p\u003e \u003cp\u003e \u003cstrong\u003eConsent to participate\u003c/strong\u003e \u003cp\u003e \u003cem\u003eInformed consent was obtained from all participants involved in the study.\u003c/em\u003e \u003c/p\u003e \u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e \u003cp\u003eThe authors received no specific funding for this study.\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eI.E.E. conceived the study, developed the theoretical framework, and led the writing of the manuscript. C.C.C. and L.O.K. contributed to the conceptual development, literature review, and critical interpretation of findings. A.G.O. and A.F.I. supported the research design, data interpretation, and methodological refinement. G.A.E. contributed to empirical analysis and contextual interpretation of organisational data. O.M.I. contributed to the ethical framing, governance analysis, and critical revisions of the manuscript. All authors reviewed, edited, and approved the final version of the manuscript and agree to be accountable for all aspects of the work.\u003c/p\u003e\u003ch2\u003eAcknowledgement\u003c/h2\u003e\u003cp\u003enone\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eDanaher, J.: Algorithmic governance: Developing a research agenda. Big Data Soc. \u003cb\u003e4\u003c/b\u003e(2), 1\u0026ndash;21 (2017). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/2053951717726554\u003c/span\u003e\u003cspan address=\"10.1177/2053951717726554\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZiewitz, M.: Governing algorithms: Myth, mess, and methods. Sci. Technol. Hum. Values. \u003cb\u003e41\u003c/b\u003e(1), 3\u0026ndash;16 (2016). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1177/0162243915608947\u003c/span\u003e\u003cspan address=\"10.1177/0162243915608947\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePasquale, F.: The black box society: The secret algorithms that control money and information. Harvard University Press (2015)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKitchin, R.: Thinking critically about and researching algorithms. Inform. Communication Soc. \u003cb\u003e20\u003c/b\u003e(1), 14\u0026ndash;29 (2017)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGillespie, T.: The relevance of algorithms. In: Gillespie, T., et al. (eds.) Media technologies, pp. 167\u0026ndash;194. MIT Press (2014)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBeer, D.: The social power of algorithms. Inform. Communication Soc. \u003cb\u003e20\u003c/b\u003e(1), 1\u0026ndash;13 (2017)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEspeland, W.N., Stevens, M.L.: Commensuration as a social process. Ann. Rev. Sociol. \u003cb\u003e24\u003c/b\u003e, 313\u0026ndash;343 (1998)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEspeland, W.N., Stevens, M.L.: A sociology of quantification. Eur. J. Sociol. \u003cb\u003e49\u003c/b\u003e(3), 401\u0026ndash;436 (2008)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePower, M.: The audit society: Rituals of verification, 2nd edn. Oxford University Press (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDesrosi\u0026egrave;res, A.: The politics of large numbers: A history of statistical reasoning. Harvard University Press (1998)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePorter, T.M.: Trust in numbers: The pursuit of objectivity in science and public life. Princeton University Press (1995)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eO\u0026rsquo;Neil, C.: Weapons of math destruction. Crown Publishing (2016)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZuboff, S.: The age of surveillance capitalism. Profile Books (2019)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAmoore, L.: Cloud ethics: Algorithms and the attributes of ourselves and others. Duke University Press (2020)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEubanks, V.: Automating inequality. St. Martin\u0026rsquo;s (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKellogg, K.C.: Subtle shifts in control: Professional autonomy and algorithmic oversight. Adm. Sci. Q. \u003cb\u003e66\u003c/b\u003e(1), 1\u0026ndash;41 (2021)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJarrahi, M.H.: Artificial intelligence and the future of work. Bus. Horiz. \u003cb\u003e61\u003c/b\u003e(4), 577\u0026ndash;586 (2018)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOrlikowski, W.J.: Sociomaterial practices. Organ. Stud. \u003cb\u003e28\u003c/b\u003e(9), 1435\u0026ndash;1448 (2007)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLeonardi, P.M.: When flexible routines meet flexible technologies. MIS Q. \u003cb\u003e35\u003c/b\u003e(1), 147\u0026ndash;167 (2011)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSuchman, L.: Human\u0026ndash;machine reconfigurations. Cambridge University Press (2007)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWinner, L.: Do artifacts have politics? Daedalus. \u003cb\u003e109\u003c/b\u003e(1), 121\u0026ndash;136 (1980)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFeenberg, A.: Transforming technology, 2nd edn. Oxford University Press (2010)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFlyvbjerg, B.: Making social science matter. Cambridge University Press (2001)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eScott, J.C.: Seeing like a state. Yale University Press (1998)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBurrell, G., Morgan, G.: Sociological paradigms and organisational analysis. Heinemann (1979)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTownley, B.: Performance appraisal and the emergence of management. Acc. Organ. Soc. \u003cb\u003e18\u003c/b\u003e(3), 221\u0026ndash;238 (1993)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRoberts, J.: The possibilities of accountability. Acc. Organ. Soc. \u003cb\u003e16\u003c/b\u003e(4), 355\u0026ndash;368 (1991)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGrey, C.: Career as a project of the self. Sociology. \u003cb\u003e28\u003c/b\u003e(2), 479\u0026ndash;497 (1994)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlvesson, M., K\u0026auml;rreman, D.: Varieties of discourse. Hum. Relat. \u003cb\u003e53\u003c/b\u003e(9), 1125\u0026ndash;1149 (2000)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWillmott, H.: Strength is ignorance; slavery is freedom. J. Manage. Stud. \u003cb\u003e30\u003c/b\u003e(4), 515\u0026ndash;552 (1993)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBoltanski, L., Chiapello, \u0026Egrave;.: The new spirit of capitalism. Verso (2005)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDourish, P.: Algorithms and their others. Big Data Soc., \u003cb\u003e3\u003c/b\u003e(2). (2016)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLatour, B.: Reassembling the social. Oxford University Press (2005)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCallon, M.: The laws of the markets. Blackwell, Oxford (1998)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBowker, G.C., Star, S.L.: Sorting things out: Classification and its consequences. MIT Press (1999)\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSelznick, P.: Leadership in administration. Harper \u0026amp; Row (1957)\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Algorithmic governance, AI ethics, Quantification and metrics, Optimisation regimes, Socio-technical systems, Responsibility and agency","lastPublishedDoi":"10.21203/rs.3.rs-8752951/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8752951/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eContemporary debates in AI ethics often associate algorithmic governance with advanced computational systems, automated decision-making, and machine learning. This focus, however, risks overlooking how similar governance effects emerge through non-AI optimisation infrastructures that precede and normalise algorithmic control. This paper introduces the concept of algorithmic governance without algorithms to examine how metrics, dashboards, and performance indicators embedded within Lean\u0026ndash;Green manufacturing systems reshape authority, agency, and ethical responsibility in organisational life. Drawing on mixed-method empirical material from Nigerian manufacturing organisations\u0026mdash;including surveys, semi-structured interviews, and observational data\u0026mdash;the study analyses how optimisation regimes function as socio-technical systems of governance rather than neutral managerial tools. The findings reveal three interrelated dynamics: the dominance of quantified metrics in decision-making, the reconfiguration of worker agency through conditional participation, and the transformation of sustainability from a collective ethical commitment into a measurable performance output. These dynamics demonstrate that ethical concerns commonly attributed to artificial intelligence\u0026mdash;such as displacement of human judgement, responsibility gaps, and moral deskilling\u0026mdash;are already present within everyday organisational technologies. By extending ethical analysis beyond narrowly defined AI systems, this paper contributes to AI ethics scholarship by highlighting the need to address governance risks embedded in pre-algorithmic optimisation infrastructures that shape how institutions value efficiency, responsibility, and human judgement.\u003c/p\u003e","manuscriptTitle":"When Metrics Govern: Pre-Algorithmic Optimisation and the Ethics of Artificial Intelligence","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-02-10 15:24:55","doi":"10.21203/rs.3.rs-8752951/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"6dddaf67-449c-4c91-a5dc-de6bfe091976","owner":[],"postedDate":"February 10th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-03-18T13:11:50+00:00","versionOfRecord":[],"versionCreatedAt":"2026-02-10 15:24:55","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8752951","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8752951","identity":"rs-8752951","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}