A Working Group on Policing and Digital Transformation
Paper 3 of 4
Tom Kirchmaier and Mick O’Connell
24 March 2026
ABSTRACT
Modern policing organisations face a structural paradox; the digital technologies now available to them are capable of radically distributing information, analytical capacity, and decision-making authority to the frontline, yet institutional architectures remain heavily hierarchical — shaped by inherited command models that more often concentrate authority at the middle and top and treat frontline officers primarily as executors rather than as empowered decision-makers. This paper examines the organisational consequences of that paradox in a digital world and argues for a fundamental reorientation of policing institutions: one that devolves decision rights, data, and tools to officers and first-line supervisors while maintaining the accountability, coordination, and standards that democratic policing requires. Drawing on organisational economics, principal-agent theory, and comparative institutional analysis, we examine the incentive structures, career pathways, training systems, compensation models, and governance arrangements through which this reorientation can be achieved. We argue that moving from rank-based pay structures toward skill-based compensation models is a necessary condition for building the human capital that digital policing demands, and that organisational resilience — in both its technological and institutional dimensions — requires the same devolved, adaptive structures that effective frontline empowerment implies.
Introduction: The Organisational Paradox of Digital Policing
Modern policing organisations face a distinctive but consequential paradox. Frontline officers possess significant discretion in their immediate interactions with the public — empowered to make swift, situational decisions that carry real authority in the moment. Yet, when it comes to shaping the organisational priorities and intermediate goals that underpin those encounters, authority remains embedded within legacy structures. Policing hierarchies, still influenced by military-style command models, often concentrate decision-making power in the middle and upper tiers. Information must travel upward before direction flows back down, producing a culture of supervision and procedural rechecking that can slow the movement of insight and initiative. As a result, while digital tools — from real-time data systems and AI analytics to mobile and sensor technologies — could, in principle, place strategic intelligence directly in the hands of officers, their full value is constrained by institutional architectures that have not evolved at the same pace.
The cost of this paradox is not merely organisational efficiency. It is, in the analysis we advance in this paper, a direct impediment to the realisation of modern consent based values in the digital age. A policing institution that cannot empower its frontline, one that cannot equip officers with the tools, data, and decision authority to act effectively in the situations they encounter, will find that the gap between technological capability and operational outcome grows rather than narrows - and will have a negative effect on public trust and confidence. Expensive systems will be procured, but their value will be dissipated by institutional structures that prevent them from being used as intended. Specialist skills will be developed, but their application will be constrained by hierarchical approval processes rooted in a model of authority that was designed for a very different operational environment.
At the same time, technological advances have dramatically widened access to information across all levels of policing, allowing supervisors and frontline officers alike to see, share, and interpret operational patterns in ways that could, in principle, democratise decision-making. Yet the formal authority structures through which such insights can be acted upon have changed far more slowly, leaving much of that potential unrealised.
This paper addresses that paradox directly. Our central argument is that digital transformation requires, and makes possible, a fundamental reorientation of policing institutions: one that devolves meaningful decision rights, data access, and analytical tools to officers and first-line supervisors, while maintaining the accountability, coordination, and standards that democratic policing requires. This is not a proposal to abolish command structures or to abandon the oversight mechanisms that hierarchical organisation provides. It is a proposal to redesign those structures so that authority is located where knowledge is greatest, that accountability travels with decision rights rather than being displaced upward, and that the human capital invested in specialist skills is allowed to generate value rather than being wasted by institutional structures that do not know what to do with it.
We develop this argument across several interconnected dimensions: the organisational economics of decision rights and information asymmetry; the design of career pathways and compensation systems that develop and retain the human capital digital policing requires; the role of first-line supervision as the critical interface between institutional authority and frontline action; the content and orientation of training; and the design of resilient systems and institutions — understood as the capacity to withstand failure, maintain legitimacy under pressure, and adapt without losing foundational commitments.
Our analysis reflects a productive tension within this working group — between the institutional experience of leaders who have navigated real-world operational constraints and the analytical perspectives of those who have studied similar challenges in other organisational contexts. The design principles we articulate in the concluding section reflect that tension honestly. They are aspirational in direction but grounded in operational reality.
The Inherited Architecture: Command Hierarchy and Its Constraints
The Military Inheritance
The command structures of modern police organisations are, in large part, originally inherited from military models that were themselves designed for a very specific operational purpose - the coordination of large numbers of individuals in conditions of physical danger, high uncertainty, and the need for rapid, disciplined collective action. In such conditions, centralised command structures offer genuine advantages. They reduce coordination costs, enable rapid response to changing circumstances, and provide clear lines of authority that prevent the confusion that decentralised decision-making can produce under pressure.
The transfer of these structures to police institutions, which were sometimes explicitly designed to be distinct from military organisations, has always been imperfect. Peel's own intention was to create a civilian institution that operated by consent rather than by command, and the analogy between the police constable / first line police officer and the military soldier was one that early reformers actively resisted. Nevertheless, the practical demands of managing large workforces engaged in activities involving physical risk, legal authority, and the potential for harm produced institutional pressures toward hierarchical organisation that proved durable. Rank structures, chain-of-command protocols, and the concentration of discretionary authority at senior levels have remained characteristic features of police institutions in most jurisdictions even as the operational environment has changed dramatically.
The result is an institutional inheritance that stands in growing tension with the operational realities of contemporary policing. The information environment in which officers operate has been transformed. Digital technologies generate vastly more data, more rapidly, than any hierarchical system can process and transmit before decisions need to be made. The nature of the decisions officers are required to make has become more complex; encounters that once called primarily for physical presence and legal authority now frequently require analytical judgment, knowledge of digital evidence, and the capacity to navigate complex multi-agency environments. And the workforce on which policing institutions depend has changed; the officers and staff now entering the profession bring educational backgrounds, technical skills, and professional expectations that were rare or absent a generation ago.
The Information Asymmetry Problem
The central organisational economics insight into the costs of hierarchical command structures concerns information asymmetry. In any large organisation, the individuals closest to the operational action, those making decisions in real time about specific situations, will typically have better information about those situations than those positioned higher in the hierarchy. When authority to make decisions is concentrated at the top of a hierarchy while information is concentrated at the bottom, two kinds of inefficiency result. First, decisions are made by those who are institutionally authorised to make them but informationally poorly placed to make them well. Second, the time and cost of transmitting information upward and instructions downward introduces delay and distortion that degrades the quality of operational response.
In the policing context, this challenge is far from theoretical. Its effects are visible in everyday practice. Officers with deep expertise in cybercrime, digital forensics, or intelligence analysis who cannot act on their professional judgment without referral to supervisors lacking the technical knowledge to assess the decisions they are asked to endorse; frontline officers who may recognise that a particular course of action is the most proportionate, lawful, or operationally sound response to a complex situation, yet remain procedurally bound to seek authorisation from individuals removed from contemporary operational realities. Specialists trained at significant institutional cost whose skills are underused because the authority to apply their judgment has not been devolved. In such contexts, what constitutes a “correct” decision is rarely absolute, it may involve trade-offs between legality, ethics, and efficiency, and this is precisely where enhanced information flow and more adaptive governance structures could enable better, more contextually balanced decision-making.
The principal-agent framework from organisational economics provides a useful analytical lens for understanding this problem. In any hierarchical structure, there is a principal, the senior authority, and an agent, the individual acting on the principal's behalf. The fundamental challenge is that the agent has information the principal lacks, and the principal cannot perfectly monitor or evaluate the agent's actions. When decision rights are concentrated with the principal, the information advantage of the agent is wasted. When decision rights are devolved to the agent, the problem of ensuring that those rights are exercised in ways consistent with the principal's objectives, accountability, must be addressed through mechanisms other than direct supervision. The design of those mechanisms is the core challenge of organisational reform in hierarchical institutions.
Hierarchy and Digital Technology: A Problematic Encounter
The introduction of digital technology into hierarchical policing organisations creates a distinctive set of challenges that compound the information asymmetry problem. Digital tools that are, in principle, capable of radically distributing analytical capacity and decision support are deployed into institutional structures that were not designed to exploit that distribution. The result is frequently that technology replicates and reinforces existing hierarchical patterns rather than disrupting them.
We identify several specific manifestations of this problem. Reporting and documentation systems designed to improve information flow frequently become compliance burdens that consume officer time without generating operational value, producing upward information flows that serve institutional accountability requirements rather than operational effectiveness. Decision-support tools that could, in principle, empower frontline judgment are configured in ways that route recommendations through approval processes, eliminating the response-time advantage that real-time decision support is supposed to provide. And the data generated by operational activity, potentially a rich resource for organisational learning and adapting management, is frequently inaccessible to the frontline officers whose experience generated it, locked in systems designed for senior management reporting rather than operational use.
The deeper problem is as much cultural as it is structural. Hierarchical organisations develop cultures that are self-reinforcing: leaders who have risen through hierarchical structures regard those structures as natural and legitimate; officers who have learned to seek approval for decisions develop habits of deference that persist even when decision rights are formally devolved and institutions that have traditionally managed risk through centralised oversight often struggle to adapt to models where accountability is maintained, but oversight is redistributed, requiring acceptance of the residual uncertainty that more distributed decision-making inevitably entails. Changing the structural architecture of policing organisations is necessary but not sufficient. The cultural conditions that make devolved decision-making work must be actively developed alongside the structural changes.
A Case for Devolved Decision-Making
Decision Rights, Information, and Accountability
The case for devolving decision rights to frontline officers and first-line supervisors rests on a convergence of arguments from organisational economics, operational experience, and democratic theory. As mentioned, the organisational economics argument is that decision rights should be located where the relevant knowledge is greatest, and that in policing, the relevant knowledge about specific operational situations is typically located at the frontline. The operational experience argument is that the complexity and pace of contemporary policing, particularly in digital domains, makes hierarchical approval processes structurally inadequate for the decisions that officers are required to make. The democratic theory argument is that officers who are empowered to exercise genuine professional judgment are also officers who can be held accountable for that judgment in ways that hierarchically deferential officers cannot.
The critical point about accountability deserves elaboration. There is a tendency to regard hierarchical structures as inherently more accountable than distributed ones, on the grounds that clear chains of command create clear lines of responsibility. But this conflates formal accountability with substantive accountability. When a decision is made by a frontline officer but formally authorised by a supervisor who lacks the expertise to evaluate it, the formal chain of responsibility runs to the supervisor while the substantive responsibility for the quality of the decision rests with the officer. The result is a system in which neither the officer nor the supervisor is genuinely accountable for the decision in any meaningful sense. The officer has fulfilled a procedural obligation by seeking approval, and the supervisor has fulfilled a formal role without the capacity for genuine oversight.
Devolving decision rights is not a call for diminished oversight but for smarter oversight. When authority sits closer to expertise, accountability becomes more meaningful. Officers entrusted to make specialist judgments can also be expected to explain, justify, and learn from them. Their decisions can be evaluated against tangible outcomes rather than compliance with procedural formality. In this sense, decentralisation does not weaken accountability, it reshapes it, aligning oversight with professional competence rather than hierarchy.
The Role of First-Line Supervision
If frontline officers are to be meaningfully empowered, the role of first-line supervision must be fundamentally reconceived. In hierarchical models, the first-line supervisor functions primarily as a conduit, passing information upward and instructions downward, and providing a layer of authorisation for decisions that officers are not trusted to make themselves. In a devolved model, the first-line supervisor becomes the critical interface between institutional authority and operational action: not a conduit but an enabler, responsible for creating the conditions in which officers can exercise empowered judgment, for providing the coaching and developmental support that effective judgment requires, and for maintaining the accountability standards that devolved decision-making demands.
We regard the quality of first-line supervision as one of the most significant determinants of both operational effectiveness and officer development in modern policing. The structural importance of the sergeant or equivalent role — the supervisor who has most direct, most frequent, and most consequential contact with the officers whose development and performance they shape - is crucial to overall success. Yet the investment in developing effective first-line supervisors is systematically inadequate in most policing institutions, and we think this is one of the most underappreciated failures in contemporary policing. The transition from officer to supervisor is frequently managed as a rank advancement rather than a role change, with insufficient attention to the different skills that effective supervisory practice requires: coaching, mentoring, performance management, ethical guidance, and the ability to create psychologically safe environments in which officers can exercise judgment and learn from mistakes.
The digital dimension of first-line supervision presents particular challenges. Supervisors whose own training and career development pre-date the digital transformation of policing may find themselves responsible for officers whose technical capabilities exceed their own, a situation that creates obvious difficulties for meaningful oversight. Addressing this requires not only investment in supervisory training that keeps pace with technological change, but also a cultural shift in the understanding of what effective supervision means: a willingness to acknowledge the expertise of officers, to exercise authority on the basis of professional judgment rather than technical knowledge, and to provide accountability structures that are appropriate to the decisions being overseen.
Maintaining Control in a Devolved System
A concern we take seriously is the question of how institutional control, coordination, and standards are maintained when authority is distributed. The concern is legitimate: policing institutions operate in contexts where errors of judgment can have severe consequences for individuals, communities, and the legitimacy of the institutions themselves. The argument for devolution is not an argument for unconstrained officer discretion, but for a more intelligent design of the boundaries within which that discretion operates.
Organisational economics offers several tools for thinking about this design challenge. Clear specification of the decisions that are devolved and those that are not, with the criteria governing that distinction made explicit and subject to revision as evidence accumulates, provides a framework for graduated devolution that extends decision rights as competence and trust develop. Performance management systems that evaluate decisions against outcomes, rather than against procedural compliance alone, create incentives for the exercise of good judgment rather than merely the avoidance of procedural error. Organisational learning systems that aggregate the lessons of distributed decision-making, identifying patterns in good and poor outcomes, and feeding those patterns back into training, guidance, and policy, enable the institutional level to benefit from the experiential knowledge generated at the frontline.
Digital technology has a crucial role to play in enabling these control mechanisms. The documentation capabilities of modern operational systems make it possible to create audit trails of decision-making that are more comprehensive and reliable than anything available in pre-digital policing. AI-enabled analysis of those trails can identify patterns that neither individual officers nor supervisors would detect, including patterns of discriminatory or disproportionate enforcement that are critical to democratic accountability. Real-time data sharing can enable coordination across distributed decision-makers in ways that hierarchical communication channels cannot match. The argument is not that technology eliminates the need for institutional control, but that it enables control mechanisms that are more sophisticated, more substantively meaningful, and less operationally costly than those available to hierarchical command structures.
Career Pathways and the Human Capital Challenge
The Structural Misalignment
One of the most consequential and consistently identified failures of contemporary policing institutions is the structural misalignment between the human capital they need and the career systems they operate. Digital policing requires deep specialist knowledge, in cybercrime investigation, digital forensics, data analysis, intelligence, online child exploitation, financial crime, and an expanding array of technical domains, that takes years to develop and that can be deployed only if institutional structures allow specialists to remain in their fields. The current structure of most policing career systems does precisely the opposite. It requires specialists to abandon their expertise in order to advance, as an example by returning to uniformed general duties as the pathway to promotion and higher pay.
The consequences of this structural misalignment are severe and cumulative. Specialists who wish to progress professionally face a choice between professional stagnation and career development at the cost of their specialism. Many choose to leave the public sector entirely, recruited aggressively by private sector organisations, particularly in technology, finance, and professional services that can offer both professional development in their specialist areas and compensation packages that policing institutions cannot match. The talent that leaves represents not only individual expertise but the institutional investment made in developing that expertise, training, mentoring, operational experience, and the tacit knowledge that accumulates over years of specialist practice. Once lost, it is difficult and expensive to replace.
We are emphatic about the scale of this problem, particularly in digital and cyber domains. The private sector's demand for officers with cybercrime investigation skills, digital forensics capability, or intelligence analysis experience is intense and growing. The compensation differential is significant, and will become more significant as the skills in question become more valuable. Policing institutions that do not redesign their career systems to offer competitive development and compensation pathways for specialists will face accelerating talent loss in precisely the domains where digital transformation makes specialist capability most critical.
Skill-Based Compensation: The Case for Reform
The core reform we advocate in the domain of career pathways and compensation is a move from rank-based pay structures — in which remuneration is determined primarily by the hierarchical position occupied — toward skill-based compensation models in which pay reflects the specialist capabilities that an individual brings to the organisation, regardless of the rank they hold. This reform addresses the structural misalignment directly: it makes it possible for a specialist to advance in professional recognition and remuneration without being required to abandon their specialism, and thereby removes the institutional incentive to leave policing that rank-based structures currently create.
The underlying argument is straightforward. In most labour markets, compensation reflects the value that individuals contribute through their capabilities, not the position they occupy in an organisational hierarchy. A surgeon earns more than a general practitioner not because of their hierarchical position but because of the specialist skills they have developed and the value those skills generate. A partner in a law firm is compensated more highly than a junior associate because their expertise and the client relationships they have built represent greater value to the firm. Policing institutions that pay officers on the basis of rank alone are applying a compensation model that is increasingly anomalous relative to the labour markets with which they compete, and doing so in a period when the specialist capabilities they need have never been more valuable.
Consider a constable / front line officer who has spent a decade developing world-class expertise in a particular domain of cybercrime investigation, or who has built irreplaceable intelligence relationships in a complex organised crime environment, creates more value for the institution, and for public safety, than many officers at a senior level like a superintendent rank. The argument for compensating that constable / front line officer accordingly, without requiring them to accept a leadership role they do not seek or to abandon the specialism that generates their value, is both economically and institutionally compelling. Models of this kind exist. Some policing systems already operate detective grade structures that allow investigators to advance through multiple levels of seniority and professional recognition while remaining in specialist roles, demonstrating that rank-independent progression is viable in practice.
We recognise that the implementation of skill-based compensation models in policing institutions faces significant institutional challenges. Pay structures are typically governed by collective agreements, and reform requires negotiation with staff associations and unions whose members have legitimate interests in the stability and fairness of existing arrangements. The assessment of specialist skills for compensation purposes requires evaluation frameworks that are robust, transparent, and resistant to manipulation — a demanding design challenge. And the cultural change required to make skill-based progression feel legitimate within institutions that have historically valorised hierarchical advancement is substantial. These are real constraints; we do not minimise them. But they are constraints that must be overcome if policing institutions are to develop and retain the human capital that digital transformation requires.
Implementing Skill-Based Pay: Navigating Staff Associations and Framework Design
As noted, transitioning from rank-based to skill-based compensation presents significant institutional hurdles, most notably the need to negotiate with staff associations and the requirement for robust assessment frameworks. Overcoming these constraints requires police leadership to shift from traditional collective bargaining postures toward collaborative, evidence-based organizational design.
Navigating Staff and Union Association and Union negotiations, Staff associations and unions have a legitimate interest in the fairness, predictability, and stability of their members' pay. To successfully negotiate the implementation of skill-based compensation, leadership must frame these reforms not as a threat to the established rank structure, but as a necessary expansion of member opportunities.
Practical strategies include:
• Focusing on Shared Interests: Leadership should emphasize that skill-based pay directly addresses member frustrations regarding stagnant career progression and inadequate compensation for specialized work. By framing the reform as a mechanism to reward officers for their expanding digital capabilities, negotiations can focus on mutual value creation.
• Utilizing Joint Fact-Finding: Rather than presenting a finished compensation model to unions, agencies should establish joint task forces—comprising HR, operational leaders, and union representatives—to cooperatively analyze private-sector salary benchmarks for critical skills (e.g., digital forensics, cyber investigation).
• Implementing "Grandfathering" and Parallel Tracks: To reduce resistance, reforms should protect the existing compensation of current personnel. Skill-based pathways should be introduced as an opt-in parallel track, allowing current officers to maintain their traditional rank-based progression while giving specialists a new avenue for advancement.
Designing Transparent and Robust Evaluation Frameworks if compensation is tied to skills rather than rank, the frameworks used to assess those skills must be completely transparent and highly resistant to manipulation. If assessments are viewed as subjective, the system will rapidly lose legitimacy.
Leadership can ensure robustness by implementing the following design principles:
• Separating Assessment from the Chain of Command: To prevent favoritism and manipulation, the evaluation of a specialist's technical skills should not be conducted solely by their direct line manager. Instead, assessments should be managed by independent, cross-functional evaluation panels or specialized HR units.
• Leveraging External Certification: Wherever possible, skill attainment should be tied to objective, industry-standard external certifications (e.g., recognized cybersecurity credentials, standardized financial investigation qualifications) rather than internal, subjective appraisals.
• Establishing Evidence-Based Portfolios: For skills that cannot be tested via standardized certification, agencies should require practitioners to submit auditable portfolios of their operational work. These portfolios can be peer-reviewed by a board of subject matter experts, similar to the credentialing processes used in the medical and legal professions.
• Instituting Regular Re-certification: Because digital technologies and cyber threats evolve rapidly, skill-based pay must not become a permanent entitlement based on a past qualification. Frameworks must require cyclical re-evaluation and continuous professional development to ensure that the compensated skills remain current and operationally valuable.
Police Staff and the Progression Gap
The career pathway challenge extends well beyond sworn officers to the civilian staff who constitute a large and growing proportion of modern policing workforces. Police staff in specialist roles, data analysts, forensic scientists, intelligence researchers, cybersecurity specialists, financial investigators, frequently hold skills and perform functions that are as operationally critical as those of sworn officers, and that are in equal or greater demand in the private sector labour market. Yet the career development and compensation opportunities available to police staff are typically far more limited than those available to sworn officers. In many organisations, there are only one or two managerial grades above the entry level of any specialist staff role, and the pay ceiling for specialist staff is often significantly lower than that of their sworn equivalents.
This situation is both unjust and operationally counterproductive. It is unjust because it creates a two-tier workforce in which the professional development opportunities available to individuals depend on a distinction, sworn versus unsworn, that is irrelevant to the specialist capabilities they bring and the value they create. It is operationally counterproductive because it channels institutional investment in staff development toward a small population of individuals who can advance managerially, leaving the majority of specialist staff without the development pathways needed to deepen their expertise and remain engaged with their work.
The argument for reform in this area parallels the argument for skill-based compensation for officers. A forensic data analyst with ten years of specialist experience in child exploitation investigations is not well served by a career structure that offers two managerial promotions as the only form of career advancement. What they need, and what the institution needs from them, is a pathway that recognises deepening expertise, provides increasing professional autonomy and recognition, and offers compensation that reflects the market value of their skills. Creating those pathways requires both the willingness to treat police staff as genuine professional partners rather than support personnel, and the institutional creativity to design career frameworks that work for roles that do not fit neatly into either uniformed police or conventional public sector employment models.
Training Content and Orientation
The reform of career pathways must be accompanied by a fundamental rethinking of training content and orientation. Existing training frameworks in most policing institutions were designed chiefly to build the competencies needed for uniformed patrol work, legal knowledge, physical skills, procedural discipline, and tactical readiness for common operational scenarios. These remain essential, but they are no longer enough. Contemporary policing increasingly demands capabilities that traditional frameworks fail to cultivate like interpreting digital evidence, navigating complex data environments, understanding the behavioural dynamics of digital offending, and analysing patterns drawn from vast information systems. Investment in such training should not be viewed solely as an institutional expense but as a long-term cost-saving measure. Retaining and developing skilled digital specialists within policing ultimately costs less than repeatedly recruiting and retraining personnel who depart after limited service, or relying on private contractors to provide critical technical expertise at a premium.
We identify a significant gap in training around criminal behaviour, not merely the legal categories and tactical indicators of specific offences, but the psychological, social, and economic dynamics that drive criminal activity in digital environments. Officers who understand why people commit cybercrimes, how online criminal networks recruit and operate, and what the behavioural signatures of digital offending look like are better equipped to prevent and investigate those offences than officers who know only the procedural steps for handling digital evidence. Developing that understanding requires training that draws on criminological and behavioural science as well as operational procedure, an investment in the intellectual as well as the practical dimensions of specialist professional development.
The digital dimension of training also raises questions about the relationship between initial training and continuous professional development. In domains where the technology and the threat landscape are changing rapidly, initial training, however comprehensive, becomes obsolete quickly. Policing institutions need to develop models of continuous professional development that keep specialist skills current as technologies evolve, that can be delivered flexibly to officers who cannot be released from operational duties for extended training periods, and that create the kind of sustained professional learning communities in which tacit expertise is shared and developed across practitioners rather than accumulated only by individuals.
Leadership, Incentives, and Organisational Culture
Incentive Architecture as Organisational Design
One of the most important and frequently underappreciated insights from organisational economics is that incentive structures are a fundamental element of organisational design, not an afterthought to be considered once structural arrangements are in place, but a primary determinant of the behaviours that organisations actually produce. The formal structures of an organisation specify what people are supposed to do; the incentive structures determine what they are likely to do. When these are misaligned, when the formal goals of an organisation are inconsistent with the incentives it creates for its members, the incentives typically win.
We are direct about this. The decisions made by police leaders — whether to invest in innovation, whether to pursue difficult reforms, whether to accept the short-term operational costs of longer-term capability development — are heavily shaped by the incentives those leaders face. A chief constable who is evaluated primarily on the basis of crime statistics, response times, and the avoidance of public controversy has incentives to prioritise operational stability over organisational reform, even when they understand intellectually that reform is necessary. A middle manager whose career advancement depends on the avoidance of visible failures has incentives to resist the structured experimentation that learning and innovation require. And an officer whose specialist expertise is valued by the private sector but undervalued by the promotional system of their employing organisation has incentives to leave rather than remain.
The reform of incentive structures must be treated as a strategic priority in its own right, not merely as a secondary outcome of organisational redesign. These reforms must extend across all levels of policing — from strategic leadership to frontline specialists — to ensure that innovation, professional growth, and organisational learning are consistently rewarded. Performance evaluation frameworks should incorporate explicit metrics that capture development and adaptability. For example, retention rates of officers with critical digital certifications, the proportion of operational decisions successfully devolved to the frontline, or measurable improvements in specialist capability and knowledge sharing. Career advancement criteria should recognise contributions to organisational development and innovation alongside operational performance, while compensation systems should create genuine incentives for retaining and deepening specialist expertise. None of these reforms will be straightforward, but without them, the structural changes proposed in this paper are unlikely to achieve lasting impact.
Culture, Psychological Safety, and the Conditions for Learning
Underlying the structural and incentive reforms we have discussed is a requirement for cultural change that is, in some respects, the most demanding element of the organisational reform agenda. The cultures of most policing institutions — shaped by decades of hierarchical command, by the occupational solidarity that derives from shared risk, and by the institutional imperative to project confidence and authority in operational contexts — are not naturally hospitable to the uncertainty, experimentation, and acknowledgment of error that learning organisations require.
Psychological safety, the shared belief among members of an organisation that they can raise concerns, acknowledge mistakes, propose novel approaches, and challenge existing practices without fear of punishment or humiliation, is a precondition for the kind of organisational learning that digital transformation requires. Officers who believe that raising concerns about the limitations of an AI system will be treated as obstructionism; supervisors who believe that acknowledging a decision error will damage their career prospects; leaders who believe that experimental approaches that fail will be used against the. These are officers, supervisors, and leaders who will not contribute to the institutional learning that enables improvement over time. The result is organisations that accumulate failures silently and repeat them, rather than learning from them openly.
Building psychological safety in policing institutions is a leadership challenge as much as a structural one. It requires leaders who model the behaviours they want to see, who acknowledge uncertainty, who discuss failures openly, who engage with criticism constructively rather than defensively. It requires performance management systems that distinguish between good process and good outcome, recognising that good decisions sometimes produce bad outcomes, and that the appropriate response to failure is learning rather than punishment. And it requires institutional cultures that treat the development of capability as a long-term investment, rather than demanding immediate returns from every initiative in ways that foreclose the iterative experimentation through which genuine learning occurs. We are clear that this cultural dimension of organisational reform is as important as any structural change — and it is, in many respects, the harder challenge.
Empowering the Frontline: A Design Framework
Distributing Data, Tools, and Decision Rights
The practical realisation of frontline empowerment requires attention to three interdependent elements: the data that officers have access to, the tools through which they can use that data, and the decision rights that determine what they can do on the basis of what they know. Progress on any one of these without corresponding progress on the others will produce limited results. Officers who have data but not the tools to analyse it cannot exploit the informational advantage that data provides. Officers who have data and tools but not the decision rights to act on their analysis must still route their judgments through approval processes that negate the operational value of real-time decision support. And officers who have decision rights but not the data and tools they need to exercise good judgment will exercise those rights on the basis of incomplete information in ways that generate poor outcomes.
Whilst it is encouraged that data flow and devolution of decision-making moves more to the frontline, this proposition does not empower those same officers to hoard such data, nor keep it out of common access to enable wider organisational use, nor does it absolve management of the responsibility of aggregating the same for its wider response to crime and criminality in a timely and agile manner.
The design challenge is therefore to create alignment across all three dimensions simultaneously, a demanding coordination task that requires institutional leadership that understands the interdependencies involved and can manage change across functions that, in most policing organisations, are managed separately. Technology procurement decisions are typically made by IT departments; training and career development decisions are made by HR functions; decision authority and governance are determined by operational command structures; and data governance is managed by information management teams. None of these functions can, in isolation, create the conditions for effective frontline empowerment. Doing so requires cross-functional coordination that policing institutions rarely manage well and that demands deliberate institutional investment.
We identify several areas where better alignment of data, tools, and decision rights could enable significant operational gains. An integrated data environment, one in which officers attending an incident can access relevant information from social services, health, prior contacts, and other sources, remains a potentially transformative capability that most forces have yet to realise, despite the technical availability of the underlying data. Achieving this, however, requires robust governance, privacy frameworks, and inter-agency agreements that ensure lawful and ethical data sharing across institutional boundaries — giving leaders confidence that integration can be achieved without compromising data protection or public trust. Similarly, the ability of specialist investigators to retrieve digital evidence without routing requests through time-consuming bureaucratic processes would remove an unnecessary operational bottleneck.
Finally, enabling first-line supervisors to monitor and support decisions through real-time data, rather than retrospective review, represents a further opportunity to elevate both operational performance and the quality of professional development.
Structured Experimentation and Organisational Learning
The reform agenda we have outlined in this paper involves significant organisational change across multiple dimensions simultaneously. Implementing such change effectively requires an approach to innovation that is both systematic and iterative, one that creates structured opportunities for experimentation, learns from the results of those experiments, and scales what works while modifying or abandoning what does not. This is the approach that successful innovation in complex organisations typically requires, and it is one that the bureaucratic and hierarchical cultures of many policing institutions make structurally difficult.
The bureaucratic character of hierarchical decision-making is, in our view, the primary inhibitor of innovative behaviour in policing organisations. When every significant departure from established practice requires approval through multiple layers of command, each of which has incentives to avoid the risk of visible failure, the institutional cost of experimentation is high and the expected institutional reward is low. The result is that innovation, when it occurs, tends to be incremental and confined to domains where approval is easy to obtain, rather than bold and targeted at the areas of greatest need.
Addressing this requires explicit institutional commitment to structured experimentation, the creation of governance arrangements that allow new approaches to be tested in defined contexts, with clear evaluation criteria and predetermined decision points at which evidence determines whether to proceed, modify, or discontinue. This is not permission for unconstrained innovation; it is a disciplined approach to managing the risks of change in ways that produce institutional learning rather than cumulative failure. Crucially, police executives must also manage the external political risks that come with experimenting in a highly visible and politically sensitive domain. Transparent governance, including clear communication of objectives, boundaries, and evaluation criteria to oversight bodies and the public, helps to distinguish responsible experimentation from recklessness. By framing pilot initiatives as evidence-gathering exercises conducted under controlled conditions, police leaders can create political space for good-faith operational mistakes without undermining public confidence. Some jurisdictions have established innovation units or technology labs within police institutions that model this approach, operating with greater autonomy from routine approval processes while maintaining robust accountability.
The key insight from these models is that agile and flexible innovation can coexist with disciplined governance and external transparency. Adoption and experimentation are separate challenges, and the latter can be managed responsibly without waiting for the former to be fully resolved.
Accountability Without Hierarchy
A concern we hear often —and take seriously — is whether meaningful accountability can be maintained in the absence of the hierarchical oversight mechanisms that current structures provide. We argue that it can, but that it requires a deliberate redesign of accountability mechanisms to reflect the devolved structure of decision-making, rather than the assumption that hierarchical oversight is the only legitimate form of accountability.
The key principle is that accountability should travel with decision rights: those who are empowered to make decisions should also be accountable for them, and the accountability mechanisms should be designed to enable genuine evaluation of the quality of decisions rather than mere procedural compliance. This requires outcome-based evaluation frameworks that assess whether decisions produced good results, not only whether they followed prescribed procedures. It requires documentation and audit systems that make the basis for individual decisions legible in retrospect — enabling evaluation, learning, and accountability without requiring real-time approval. And it requires independent oversight mechanisms that can evaluate patterns of decision-making across the organization, identifying systemic problems, discriminatory trends, or capability gaps that individual supervisors might miss.
Digital technology is, again, a crucial enabler of this redesigned accountability model. The documentation capabilities of digital operational systems make it possible to create audit trails that are more comprehensive and more reliable than anything that manual recording systems could produce. AI-enabled pattern analysis of those audit trails can identify systemic issues that neither officers nor supervisors would notice in individual cases. And the real-time monitoring capabilities of networked operational systems can enable supervisors to provide developmental support to officers in ways that are both more immediate and less intrusive than traditional oversight models. The argument is that digital tools make it possible to have more accountability with less hierarchy — a genuinely attractive proposition for policing institutions that are trying to balance democratic accountability with operational effectiveness.
Design Principles for Organisational Reform
We propose the following design principles for policing organisations seeking to address the structural challenges identified in this paper. These principles build on those introduced in our earlier work and are intended as normative foundations for institutional design rather than prescriptive blueprints.
Importantly, their implementation should follow a phased and context-sensitive approach, recognising that organisational silos, legacy systems, and local operational cultures make uniform reform neither feasible nor desirable.
As an indicative sequence:
Phase 1 should focus on establishing dedicated innovation units or technology labs to pilot and evaluate new approaches within controlled environments;
Phase 2 should focus on redefining the training and roles of first-line supervisors, the critical interface for frontline empowerment;
Phase 3 should introduce skill-based compensation pathways, beginning with high-demand, high-flight-risk domains such as digital forensics and cybercrime investigation.
The Design Principles
Locate Decision Rights Where Knowledge Is Greatest
Decision authority should be devolved to the level at which the relevant knowledge for a given category of decision is most reliably available — with exceptions for decisions whose consequences are sufficiently significant, irreversible, or contested that the accountability value of senior authorisation justifies the operational cost of referral. The baseline assumption should be empowerment rather than restriction; the burden of justification should fall on those who would concentrate decision rights upward, not on those who would devolve them downward.
Invest in First-Line Supervision as the Critical Interface
The quality of first-line supervision is a primary determinant of operational effectiveness, officer development, and the practical realisation of accountability in devolved systems. Investment in developing effective supervisors through training, coaching, performance management, and the deliberate design of supervisory roles should be treated as a strategic priority, not an administrative overhead. The transition from officer to supervisor should be managed as a role change requiring new skills and orientations, not merely as a rank advancement.
Align Compensation with Value, Not Rank
Skill-based compensation models that allow specialist expertise to be rewarded appropriately, regardless of hierarchical position, are a necessary condition for developing and retaining the human capital that digital policing requires. The design and implementation of such models, including the evaluation frameworks needed to assess specialist capabilities, and the negotiation processes needed to secure their adoption within existing employment frameworks, should be treated as an urgent organisational priority.
Design Career Pathways for Both Officers and Staff
Career development and progression opportunities should be designed for the full range of specialists, sworn and unsworn, that modern policing requires, on the basis of the value they create rather than the formal category they occupy. Police staff who hold critical specialist skills should have access to career pathways that enable genuine professional development and appropriate professional recognition, rather than being limited to the one or two managerial grades that most current structures provide.
Build Accountability into the Architecture of Devolved Systems
Accountability mechanisms should be redesigned to match the devolved character of decision-making, rather than maintained in hierarchical forms that are no longer suited to the decisions being made. This requires outcome-based evaluation frameworks, digital audit and documentation systems that enable genuine retrospective review, and independent oversight mechanisms capable of identifying systemic patterns that individual supervisors would miss.
Treat Resilience as a Design Principle, Not a Reactive Capacity
Resilience comes in two distinct but deeply connected forms, and both must be designed in from the outset rather than bolted on after failures occur.
Technological resilience, the capacity of digital systems to maintain functionality and integrity under stress, attack, or failure, is too often treated as the exclusive concern of IT departments rather than a strategic leadership priority. It is both. Police organisations hold some of the most sensitive data in the public sector and are attractive targets for adversarial actors. Legacy systems that are outdated, poorly documented, or dependent on knowledge held by a handful of individuals nearing retirement are inherently fragile. Cloud dependencies can transfer resilience risk to private providers whose interests do not always align with those of a public institution. And the expanding networks of data sharing and interoperability that digital transformation creates generate new attack surfaces that must be actively managed. Critically, when significant incidents occur, the failures that matter most are rarely technical, they are failures of human preparation, unclear incident protocols, and leadership that was not ready to manage the response. Technological resilience is therefore as much an organisational and cultural challenge as a technical one.
Institutional resilience, the capacity of policing organisations to maintain legitimacy and effectiveness through crisis and change, is inseparable from the broader empowerment agenda this paper advances. Organisations that concentrate all consequential decisions at the top are brittle. When the command centre is disrupted, by a cyber attack, a leadership crisis, or a public scandal, they lose the ability to function in ways that distributed organisations do not. Resilience and empowerment are complementary, not competing. An organisation that has genuinely devolved decision-making has, by design, built a more resilient architecture. Equally, a policing institution that has invested in public trust through transparency and accountability carries a reserve of legitimacy that enables it to navigate crises without catastrophic loss of public confidence, whereas institutions that have treated trust-building as secondary find, precisely when it matters most, that the consent they depend on has eroded.
In both domains, the principle is the same: resilience must be anticipated, designed for, and regularly tested; not improvised in the aftermath of failure.
Create Conditions for Continuous Learning
Organisational learning requires cultures of psychological safety, structured opportunities for experimentation, and feedback loops that connect the lessons of distributed decision-making to institutional policy and practice. These conditions must be actively built and maintained by institutional leadership. They do not emerge spontaneously in organisations shaped by hierarchical command cultures and the imperative to project operational confidence.
Conclusion: From Inherited Hierarchy to Adaptive Organisation
The policing organisations of the digital age face a structural challenge that no amount of technological investment can resolve without accompanying organisational reform. The hierarchical command structures inherited from military models, and sustained by institutional cultures that have been shaped by those structures over decades, are not adequate to the demands of an operational environment characterised by distributed information, rapid change, specialist knowledge, and the need for adaptive, empowered frontline response.
The case for change is not merely one of operational efficiency, though the efficiency gains from getting this right are substantial. It is a case grounded in the same democratic values that consent based principles and Peelian policing models espouse. Institutions that concentrate authority at the top and treat frontline officers as executors rather than decision-makers are not only institutionally inefficient. They are democratic accountability failures. The officer who can only act with approval is not an officer who can be held accountable for the quality of their judgment. The institution that manages risk by centralising authority is not an institution that has solved the accountability problem — it has displaced it upward to individuals who lack the information and expertise to exercise genuine oversight.
The reforms we have argued for in this paper are demanding. They require changes to compensation systems, career pathways, training content, supervisory practice, governance arrangements, resilience planning, and institutional culture — simultaneously and in ways that are mutually reinforcing rather than sequentially and in isolation. They require leaders who understand the organisational economics of their own institutions and who are willing to accept the short-term costs of change in pursuit of long-term capability. And they require the kind of sustained institutional commitment that is difficult to maintain in the face of the operational pressures and political contingencies that policing leaders manage every day.
We do not minimise these challenges. But we note, as a working group that includes among its members senior leaders who have implemented significant organisational changes in large and complex policing institutions, that these reforms are not only theoretically compelling, they are practically achievable. There is evidence from other policing agencies and jurisdictions that elements of the reforms proposed here — including devolved decision-making, skill-based career development, and deliberate cultivation of organisational resilience — can yield improved operational performance, stronger institutional capacity, and enhanced democratic accountability. The question is not whether the reforms are possible. It is whether the institutions that need them most will find the leadership courage and the institutional capacity to pursue them.
The next paper in this series will address the technological infrastructure dimension of the reform agenda — the questions of platform ownership, data architecture, AI governance, and interoperability that must be resolved alongside the organisational changes we have examined here. Together, the three papers constitute a framework for understanding what legitimate and effective policing in the digital age requires. We offer them, as we have said throughout, not as settled conclusions but as contributions to the ongoing dialogue through which that understanding will be developed.
About This Initiative
This is the third in a series of four publications emerging from our working group, convened at the Centre for Economic Performance, London School of Economics and Political Science, in February 2026. We bring together senior police leaders, academics, technologists, and other practitioners to develop a comprehensive framework for understanding and guiding the future of policing's contribution to public safety provision in the digital era.
The initiative is deliberately locationally agnostic, emphasising questions of organisational design, digitalisation, and interoperability that transcend particular jurisdictional or ideological contexts. It is positioned as an ongoing working group rather than a discrete conference, with plans to continuously revisit and refine analyses based on feedback and emerging developments. This approach treats each paper as a contribution to ongoing dialogue rather than a definitive pronouncement, reflecting the genuine uncertainty and rapid evolution that characterise the landscape under analysis.
The central research question animating the initiative asks: “What institutional forms, organisational structures, and technological systems are required for legitimate and effective policing in an increasingly digital society?” The three thematic papers in the series address this question from complementary perspectives: this first paper establishes the normative foundations; subsequent papers address organisational design and technological infrastructure respectively. Together, they are intended to inform and shape early discourse between the public and policing institutions toward a renewed consent based and somewhat Peelian style consensus of public trust in policing approaches for the twenty-first century.
We invite feedback, critique, and engagement with these ideas as part of an ongoing dialogue about the future of democratic policing in an age of digital transformation. Please post your comments on our discussion page. Or if you would like to submit your comments privately or anonymously please email them to T.Kirchmaier@lse.ac.uk.
Participants
Irakli Beridze, UNICRI, United Nations
Chris Church, INTERPOL
Megan Hoey, Capgemini
Tom Kirchmaier, CEP/LSE
Rachel Lewis, City of London Police
William Lyne, Met Police
João Mota, VOID Software
Mick O'Connell, UNICRI & Critical Insights Consultancy Ltd
Emily Owens, CEP/LSE & UCI
Emma Persson, UNICRI, United Nations, Centre for AI and Robotics
Michalis Pittalis, Cyprus Police
Liam Price, Royal Canadian Mounted Police (RCMP)
Dominic Reese, North Rhine-Westphalia Police, Germany
Inger Marie Sunde, Politihøgskolen / Norwegian Police University College
Chris Sykes, Greater Manchester Police (GMP)
Dick van Veldhuizen, Roseman Labs
Ben Waites, Europol
Liz Ward, Met Police
The views expressed in this paper are those of the authors and do not necessarily reflect the official positions or policies of the organizations with which they are affiliated. Special thanks to Catherine Ojo and Janey Tietz for invaluable support with this project. The proceedings were transcribed using Notion, the summaries were produced using Notion and Claude, and the presentation drafted using Gamma. All automated summaries were edited and signed off by us humans. All errors are our own.
24 March 2026