Public Administration Systems and Models

Why this matters in public value

Most leaders did not choose the administrative model they run. They inherited a civil service grading structure built for a hierarchy, a portfolio of outsourced contracts built for a market, a set of partnership boards built for a network, and a handful of digital services bolted onto all three — often in the same organization, sometimes in the same directorate. Each layer was a reasonable answer to the reform agenda of its decade, and each layer still shapes what the organization can do today, whether or not anyone currently working there chose it.

This matters because the four paradigms compared in this topic — Traditional Public Administration, New Public Management, New Public Governance, and digital-era governance — are not simply successive fashions that each replaced the last. They are different answers to a permanent design question: should coordination run through rules and hierarchy, through price and competition, through negotiated relationships, or through shared digital infrastructure? Every live organization is a blend, and the blend is rarely the product of a single deliberate decision. A leader who cannot name which paradigm governs a given function is not neutral between the four; they are simply unaware of which one is quietly in charge.

The stakes are practical, not academic. A service redesigned around market competition inherits NPM's strengths — sharper cost discipline, clearer output specification — and its weaknesses — fragmented accountability, an assumption that citizens behave like consumers. A service redesigned around collaborative networks inherits New Public Governance's strengths — genuine co-production, resilience through relationships — and its weaknesses — slower decision-making, blurred lines of formal accountability. Choosing badly, or drifting into a model by accident, costs public value for years before anyone traces the cause back to a structural choice made, or defaulted into, long before.

Core concepts

Public administration, as a field, is both the practice of implementing public policy through organizations and the academic discipline that studies how that implementation works. This topic is the home topic for the typology that structures the whole of Part 3: the major paradigms that describe how public organizations have been designed and reformed, in rough historical sequence but with all four still coexisting in practice today.

Traditional Public Administration (TPA) is the paradigm built on bureaucracy in the technical, non-pejorative sense the German sociologist Max Weber gave the term. Weber described an "ideal type" — an analytical model, not a claim that any real organization matches it perfectly — with a fixed hierarchy of offices, decisions made by applying written rules rather than personal discretion, a clear separation between an official's public role and private interests, appointment and promotion by demonstrated competence rather than patronage, and a permanent, salaried civil service whose members build careers inside the institution rather than moving in and out with each election. TPA also rests on the politics–administration dichotomy: elected or appointed leaders set policy, and a neutral, professional bureaucracy carries it out impartially regardless of who is in office. Its strengths are real and durable — predictability, equal treatment under published rules, resistance to corruption and favouritism, and institutional memory that survives political turnover. Its characteristic failure is that rule-following becomes an end in itself: proper procedure substitutes for the outcome the procedure was meant to secure, a pattern often called red tape or, more formally, goal displacement.

New Public Management is the reform wave that dominated public-sector thinking from the early 1980s into the 2000s, first in the United Kingdom, New Zealand, and Australia and then more widely. The scholar Christopher Hood, who named and codified the doctrine in 1991, identified its recurring components: breaking up large bureaucracies into smaller, single-purpose units; introducing competition through contracting-out and internal markets; borrowing private-sector management styles and giving managers more discretion in exchange for holding them accountable to explicit, measurable targets; and judging public organizations by the efficiency and economy of their outputs rather than only the propriety of their processes. NPM treated the citizen primarily as a customer and the public manager primarily as an entrepreneur constrained by a contract rather than a rule book. Where it worked, it cut genuine waste and sharpened services that had drifted into producer-driven complacency. Where it failed, it fragmented accountability across many small contracted units, encouraged gaming of narrow performance indicators rather than genuine improvement (see Topic 3.11 — Quality, Safety, and Performance Management for Goodhart's law in this context), and treated public value as if it collapsed onto the single dimension NPM could most easily measure: cost per unit of output.

New Public Governance (NPG) is the paradigm, associated most closely with the scholar Stephen Osborne, that treats public value as produced through networks of interdependent actors — government agencies, contracted providers, nonprofits, and citizens themselves — rather than through hierarchy or market competition alone. Where TPA coordinates through rules and NPM coordinates through price and contract, NPG coordinates through negotiated, often informal relationships, described in the literature as network governance. Its central mechanism is co-production: citizens and communities are not only recipients of services but active participants in designing and delivering them, a shift this book treats in full elsewhere (see Topic 1.5 — Stakeholders, Citizens, and Co-Production). NPG's strength is that it takes seriously what TPA and NPM both tend to miss — that many public problems (crime, chronic illness, community resilience) are only partly solvable by any single organization acting alone, however well designed. Its characteristic failure is diffusion: when everyone is a partner, accountability for a bad outcome can become genuinely difficult to locate.

Digital-era governance (DEG) is the pattern the political scientists Patrick Dunleavy, Helen Margetts, Simon Bastow, and Jane Tinkler identified in 2006 as NPM's successor in the countries that had pursued it furthest: reintegrating the fragmented, contracted-out functions NPM had split apart, redesigning services around citizens' actual needs rather than agency silos, and using digital technology not merely to automate existing paper processes but to restructure how government itself is organized. This topic places digital-era governance in the typology; how it plays out in practice — digital-first service design, shared platforms, "government as a platform" — is covered in depth elsewhere (see Topic 5.1 — Digital Government and Digital Transformation).

Public value theory as a response to NPM's limits. Mark Moore's strategic triangle, the framework this book is built on, did not emerge in a vacuum. Moore developed it in the early 1990s partly as a corrective to NPM's tendency to judge public organizations chiefly on efficiency and measurable output, arguing that a public manager's job is not only to run an efficient operation but to build legitimacy and support for a value proposition that a market test alone cannot validate (see Topic 1.2 — The Strategic Triangle: Legitimacy, Value, and Capacity). Gerry Stoker later described "public value management" explicitly as a successor narrative to NPM, one that restores questions of democratic legitimacy and collective preference that a purely managerial, market-facing model tends to crowd out. Public value theory does not reject NPM's insights about efficiency, nor NPG's insights about networks, nor DEG's insights about digital reintegration; it insists that all of them be judged against a broader test — whether the result is something citizens, collectively, would authorize spending shared resources to produce.

Best practices

1. Name the paradigm actually governing each major function, not the one your organization chart implies. A directorate can look hierarchical on paper while running almost entirely on contracted-out delivery; find out which model is really coordinating the work before redesigning anything.
2. Match the paradigm to the problem, rather than applying one model everywhere out of habit. Routine, high-volume, rule-bound transactions often suit TPA's predictability; genuinely novel or cross-cutting problems often need NPG's networked coordination.
3. Treat contracting-out as a coordination choice with real trade-offs, not a default cost-saving move. Before extending a market-style solution, ask explicitly what accountability, institutional memory, and coordination capacity leave the organization along with the function.
4. Specify performance targets narrowly enough to be measurable and broadly enough to avoid distorting the behavior they are meant to improve. A target that can be hit without improving the underlying outcome will eventually be hit that way (see Topic 3.11 — Quality, Safety, and Performance Management).
5. Invest in relationship management as a distinct discipline when working through networks. New Public Governance succeeds or fails on the quality of ongoing relationships between partners, which needs dedicated time and skill, not an assumption that goodwill alone will sustain it.
6. Preserve a career civil service core even inside a heavily contracted-out system. Retaining enough in-house expertise to specify, oversee, and if necessary take back a contracted function protects the organization from becoming a client that cannot judge its own supplier's performance.
7. Reintegrate fragmented services around the citizen's need, not the agency's organization chart, when digital redesign creates the opportunity. Digital-era governance's core insight is that technology can undo the fragmentation NPM's disaggregation left behind, but only if redesign starts from the user's journey (see Topic 5.1 — Digital Government and Digital Transformation).
8. Audit which paradigm's failure mode your organization is currently most exposed to. TPA drifts toward rigidity, NPM toward gaming and fragmentation, NPG toward diffused accountability, and DEG toward technology-led change that outruns institutional capacity; know which risk is live now.
9. Use co-production deliberately where it fits, rather than treating citizen involvement as consultation theater. Genuine co-production changes who does the work, not only who is asked their opinion about it (see Topic 1.5 — Stakeholders, Citizens, and Co-Production).
10. Build in an explicit accountability map whenever a function spans more than one paradigm. A service delivered by a contracted provider, coordinated through a partnership board, and accessed through a shared digital platform needs one place where a citizen or an auditor can find out who answers for what.
11. Revisit administrative design choices on a fixed cycle, not only after a visible failure. Models that suited yesterday's problem and yesterday's technology quietly become liabilities; a scheduled review catches the drift before a crisis forces it.
12. Teach the typology to staff below senior leadership. Frontline and middle managers who can name which paradigm they are operating in make better local judgment calls than those who experience the blend only as unexplained inconsistency.

Questions to discuss with your team

Which of the four paradigms — hierarchy, market, network, or platform — actually coordinates our most important service, and does everyone on this team agree? Many organizations discover, when they ask this question directly, that different senior leaders hold different mental models of how their own service is coordinated. Use the discussion to test whether the formal structure, the budget mechanism, and the accountability relationships all point to the same paradigm or whether they have quietly drifted apart. Where they disagree, that mismatch is often the root of a recurring operational frustration nobody has previously traced to its structural cause. Treat this as a diagnostic question, not a rhetorical one — write down the answer and revisit it in a year.

Where has contracting-out or market-style competition left us with a coordination or accountability gap we have not named? New Public Management's disaggregation was often the right call at the time, but every disaggregation creates a seam, and seams are where accountability gets lost. Use this question to map every major contracted function and ask who is accountable, to whom, when the contracted provider underperforms. Discuss whether the organization retains enough in-house capability to notice underperformance before a citizen or an auditor does. This is not an argument for reversing every contract, only for knowing precisely where the seams are.

Which of our current performance targets could be hit without the underlying outcome actually improving? This is Goodhart's law in practice, and almost every organization that has adopted numeric targets is carrying at least one example. Use this question to walk through your published or internal targets one at a time and ask, honestly, what behavior each one rewards. Discuss whether any target has already begun to distort frontline behavior, and what a better proxy measure would look like. Treat a defensive answer — "nobody would actually game that" — as itself worth investigating.

Where do we depend on a network of partners, and do we manage those relationships with the same rigor we apply to a formal contract? New Public Governance's central claim is that some problems can only be solved through sustained, negotiated relationships across organizational boundaries, but relationships without discipline drift into either neglect or dependency. Use this question to list your genuinely load-bearing partnerships and check when each was last actively reviewed, rather than simply assumed to be working. Discuss what would happen to service continuity if any one partner withdrew, and whether that risk has been named anywhere formally.

If we redesigned this service today, digital-first and around the citizen's actual need, what would we stop doing? Digital-era governance's promise is reintegration — undoing the fragmentation earlier reform waves left behind — but realizing it usually means retiring something, not only adding a new digital front end onto an unchanged back office. Use this question to identify one service where the current design still reflects a paradigm the organization has otherwise moved on from. Discuss what capability, contract, or internal boundary would need to change to deliver the redesign in practice, not only on a diagram.

Which paradigm's characteristic failure mode are we currently most exposed to, and who owns fixing it? Rigidity for hierarchy, gaming and fragmentation for markets, diffused accountability for networks, and change that outruns institutional readiness for digital reintegration are not equally distributed risks at any one time; one is usually live right now. Use this question to name the organization's current exposure honestly, rather than defaulting to whichever failure mode is most discussed in the sector generally. Discuss who has the authority and the resource to address it, and by when.

In practice: a public value example

The vehicle-licensing agency of a mid-sized country — fictional, illustrative of a pattern rather than a report on any real body — had been built in the 1970s as a textbook example of Traditional Public Administration: a single hierarchical department, staffed by a career civil service, processing license applications against a published rule book that had barely changed in two decades. Processing times were consistent and corruption was essentially unknown, but the agency was widely seen as slow, and a 1990s reform program, following the New Public Management wave sweeping the wider civil service, contracted out the agency's back-office data processing to three competing private providers, paid per application processed, while keeping policy and enforcement in-house.

The contracting-out achieved its narrow goal: unit costs fell and processing volumes rose. But a new problem emerged that the original design had not anticipated. Applications that fell outside the standard categories — a vehicle imported under an unusual customs arrangement, an applicant with a disability requiring an adapted license process — bounced between the three contracted providers and the in-house policy team, each of which had a narrow, separately incentivized remit and no shared case-ownership. Citizens with straightforward applications were well served; citizens with anything atypical experienced weeks of unexplained delay, and the agency's own staff could not always say which provider currently held a given file.

Rather than reversing the contracts outright, the agency's leadership, newly arrived from a background in service design, treated the problem as a coordination-model mismatch rather than a vendor-performance issue. They convened the three providers and the in-house policy team as a standing network — a New Public Governance-style arrangement, with a shared case-tracking system and a jointly owned service-level commitment for atypical applications specifically, rather than only for the standard high-volume ones the original contracts had specified. Alongside this, they commissioned a single digital front end that let an applicant track their case regardless of which provider currently held it, reintegrating the citizen's experience even though the back-office functions remained split across separate organizations — a digital-era governance move layered onto an NPG-style network, both built on top of an NPM-era contracting structure that nobody proposed to fully unwind.

Two years on, standard applications remained as fast and cheap as the contracting model had made them; atypical applications, previously the agency's most visible failure, fell from a median of six weeks to under two. The lesson the agency's leadership drew was not that any one paradigm had been wrong, but that a service built by successive reform waves needed someone to actively coordinate across the seams those waves had left behind — work that none of the original models, pursued in isolation, was designed to do.

Four sector lenses

Local government

Local authorities typically run the widest blend of all four paradigms inside a single organization, often within a single service — a Traditional Public Administration core for regulatory functions such as planning permission, NPM-style contracted-out delivery for waste collection or leisure services, NPG-style partnership boards for community safety or public health, and a growing DEG layer of shared digital platforms for resident-facing transactions. This proximity to residents makes paradigm mismatches highly visible quickly: a resident does not experience "a network governance failure", only a confusing runaround between departments and contractors, so local leaders often become the first to notice a seam that a national-level review would take years to surface.

National government

National administrations carry the deepest institutional memory of each reform wave, because each wave typically began as a central-government policy before spreading outward, and the resulting layering is correspondingly thick: a permanent civil service built on TPA principles, executive agencies and contracted delivery built on NPM principles, cross-departmental units built on NPG principles to tackle problems no single ministry owns, and a central digital or "government as a platform" function built on DEG principles, often competing with departmental IT for authority and budget (see Topic 5.1 — Digital Government and Digital Transformation). The politics–administration dichotomy at the heart of TPA is also tested hardest here, since national civil servants serve ministers whose electoral incentives can pull against the neutral, rule-bound implementation the traditional model assumes.

Social sector and nonprofit

Nonprofits and NGOs sit structurally closer to New Public Governance than any of the other three paradigms, because their funding and delivery model is usually already a network — grants and contracts from government, philanthropic funders, and often volunteer or community co-production woven through service delivery from the outset (see Topic 1.5 — Stakeholders, Citizens, and Co-Production). Where the sector encounters NPM most directly is as a subordinate party to it: many nonprofits deliver services as competitively tendered government contractors, importing NPM's performance-target discipline into an organization whose own governance culture and funding base may be built on very different assumptions about mission and relationship.

Multilateral and international

Multilateral bodies operate almost entirely through New Public Governance's logic by necessity, since they have no hierarchy to command member states and no market to enforce compliance — only negotiated relationships, shared standards, and each member's voluntary cooperation. Where Traditional Public Administration and New Public Management both assume a single accountable organization behind a decision, a multilateral program's accountability is distributed across donor states, recipient states, implementing partners, and the multilateral secretariat itself, making the "stealth mandate" and diffused-accountability failure modes of NPG especially persistent at this scale. Digital-era reintegration is correspondingly harder here too, since "the citizen's journey" spans jurisdictions with no shared digital infrastructure to reintegrate it around.

Common failure modes

Paradigm drift. An organization changes its coordination model piecemeal, one contract or one partnership at a time, without anyone ever deciding to change the model as a whole. The cost is not a single failure but a slow accumulation of internal inconsistency — different parts of the same organization running on incompatible assumptions about how decisions get made and who is accountable for what.

Contracting without retained capability. An organization contracts out a function under NPM logic and, over time, loses the in-house expertise needed to specify, monitor, or take back that function if the contract fails. The cost surfaces expensively and suddenly, usually at contract renewal or provider failure, when the organization discovers it can no longer credibly judge its own supplier's performance.

Target gaming. Under performance-target regimes inherited from NPM, staff and providers optimize for the measured proxy rather than the outcome it was meant to represent — a pattern captured by Goodhart's law (see Topic 3.11 — Quality, Safety, and Performance Management). The cost is a service that looks successful in its own reporting while the underlying problem it exists to solve persists or worsens.

Partnership without accountability. A New Public Governance-style network is convened to solve a cross-cutting problem, but no single member is ever made accountable for the network's collective failure, so when the outcome disappoints, each partner can truthfully say the failure was not solely theirs. The cost is that genuinely hard, cross-cutting problems attract the least accountable form of organization precisely because they need it most.

Digital veneer over unreformed process. A digital front end is added to a service under the banner of digital-era governance without reintegrating the fragmented back-office processes it sits in front of. The cost is a service that looks modern to a citizen at first contact and behaves exactly as fragmented as before the moment a case falls outside the routine path — the pattern behind the worked example above.

Maturity model

Checklist

• Map which of the four paradigms — hierarchy, market, network, platform — actually coordinates each major function.
• Identify every contracted-out function and confirm retained in-house capability to specify and monitor it.
• Test each published performance target for a plausible gaming strategy before or during use.
• List load-bearing external partnerships and confirm each has an active, recently reviewed relationship owner.
• Name, in writing, where accountability for a cross-boundary or cross-partner failure currently sits.
• Identify any digital front end added without reintegrating the back-office process behind it.
• Check whether the career civil service core has the depth to survive turnover in contracted or partner organizations.
• Confirm frontline and middle managers can name the paradigm governing their own service.
• Schedule a fixed-cycle review of administrative design choices, not only a post-failure review.
• Identify which paradigm's characteristic failure mode the organization is currently most exposed to.
• Confirm co-production, where used, changes who does the work and not only who is consulted.

Key sources

• Max Weber's theory of bureaucracy, set out in Economy and Society (published posthumously, 1922) — the foundational account of Traditional Public Administration's ideal type.
• Christopher Hood, "A Public Management for All Seasons?", Public Administration, 69 (1991) — the paper that named and codified New Public Management.
• Stephen P. Osborne, "The New Public Governance?", Public Management Review (2006) — the founding statement of New Public Governance as a distinct paradigm.
• Patrick Dunleavy, Helen Margetts, Simon Bastow and Jane Tinkler, "New Public Management Is Dead—Long Live Digital-Era Governance", Journal of Public Administration Research and Theory (2006) — the source of the digital-era governance concept.
• Mark H. Moore, Creating Public Value: Strategic Management in Government (1995), and Gerry Stoker's work on public value management as a successor narrative to NPM.
• OECD work on public governance reform and Government at a Glance, comparing administrative arrangements across member countries.

References

1. Public administration — Wikipedia — https://en.wikipedia.org/wiki/Public_administration
2. New Public Management — Wikipedia — https://en.wikipedia.org/wiki/New_Public_Management
3. Bureaucracy — Wikipedia — https://en.wikipedia.org/wiki/Bureaucracy
4. Max Weber — Wikipedia — https://en.wikipedia.org/wiki/Max_Weber
5. Civil service — Wikipedia — https://en.wikipedia.org/wiki/Civil_service
6. Christopher Hood — Wikipedia — https://en.wikipedia.org/wiki/Christopher_Hood
7. Outsourcing — Wikipedia — https://en.wikipedia.org/wiki/Outsourcing
8. Performance indicator — Wikipedia — https://en.wikipedia.org/wiki/Performance_indicator
9. Network governance — Wikipedia — https://en.wikipedia.org/wiki/Network_governance
10. Co-production (approach) — Wikipedia — https://en.wikipedia.org/wiki/Co-production_(public_services)
11. Digital era governance — Wikipedia — https://en.wikipedia.org/wiki/Digital_era_governance
12. Max Weber, Economy and Society — theory of bureaucracy (secondary discussion) — Stanford Encyclopedia of Philosophy — https://plato.stanford.edu/entries/weber/
13. Christopher Hood, "A Public Management for All Seasons?" — Public Administration / Wiley — https://doi.org/10.1111/j.1467-9299.1991.tb00779.x
14. Stephen P. Osborne, "The New Public Governance?" — Public Management Review / Taylor & Francis — https://doi.org/10.1080/14719030600853022
15. Patrick Dunleavy, Helen Margetts, Simon Bastow and Jane Tinkler, "New Public Management Is Dead—Long Live Digital-Era Governance" — Journal of Public Administration Research and Theory / Oxford University Press — https://doi.org/10.1093/jopart/mui057
16. Governance — OECD — https://www.oecd.org/en/topics/governance.html
17. Governance topic page — World Bank — https://www.worldbank.org/en/topic/governance