Innovation and Public Entrepreneurship

Why this matters in public value

Most public and social-sector organizations can point to at least one small team, tucked into a corner of the organization chart, that is allowed to work differently: an innovation lab, a digital unit, a "skunkworks" program, a foundation-funded pilot. These units exist because routine public administration is, by design, cautious — built on precedent, audit trails, and equal treatment, all of which make it slow to try something genuinely new (see Topic 3.1 — Public Administration Systems and Models for why Traditional Public Administration's strengths and its caution are the same feature). Somebody still has to originate the new idea, test it cheaply, and carry it through the organization's own skepticism, and that is entrepreneurial work whether or not anyone gives it that name.

The harder problem is not starting; it is what happens next. A city pilots an on-demand transport scheme, a housing department redesigns its intake process, a nonprofit tests a new case-management model — and the pilot works, gets written up, wins an award, and then largely stays where it was born. The organization that did not build it rarely adopts it wholesale, even when the evidence is good and the need is identical. That gap between a good pilot and its spread is not mainly a communication failure; it is a mismatch of legal authority, political conditions, and delivery capacity between the place that innovated and the place being asked to follow.

The stakes are real because public failure is not private failure. A start-up that fails quietly returns its investors' capital and closes; a public pilot that fails visibly can cost an elected official their mandate, a permanent secretary their credibility, and the next reformer their room to try again. An organization that never protects space for experimentation stagnates behind problems it cannot address with existing tools. An organization that protects that space without discipline burns legitimacy on pilots that were never going to satisfy citizens, and both failures are avoidable with the same underlying method: treat innovation as a strategy to be tested, not an act of faith to be funded.

Core concepts

Innovation, in the general sense this topic starts from, is the practical implementation of an idea that creates or improves a good, service, or process — distinct from mere invention, which stops at the idea itself (see innovation). In the public sector specifically, innovation usually means a new way of defining a problem, designing a service, or delivering an outcome that the organization was not previously using, whether or not the underlying idea is new to the world.

A public-sector innovation lab is a dedicated team, unit, or fund — inside government, attached to it, or funded alongside it by a foundation — whose explicit purpose is to structure and embed innovation methods, such as service design, data analysis, prototyping, and rapid testing, in the pursuit of a public or social problem. The British innovation foundation Nesta (charity), which has run, designed, and studied such units across many countries, describes them as having moved "from margin to mainstream" over roughly the past decade, with governments increasingly institutionalizing what used to be one-off pilot projects into standing units. What distinguishes a lab from an ordinary policy or delivery team is partial insulation: a lab is usually given explicit permission, a separate budget line, or a protected mandate that lets it take on more delivery risk than a line department normally could, precisely because the rest of the organization is optimized for consistency and audit, not for trying ten things and expecting several to fail (see Topic 3.10 — Regulation and Public Risk Management for how risk appetite is set and governed more generally). A lab's insulation is deliberately partial, not absolute — it still answers, eventually, to the same legitimacy and accountability structures as the rest of the organization; a lab that is insulated from accountability rather than merely from routine process is not protected innovation, it is an ungoverned parallel bureaucracy.

Public entrepreneurship is the application of an entrepreneurial disposition — noticing an opportunity nobody else has claimed, working out how to capture it, and then doing so — to the pursuit of public value rather than private profit. The term and its clearest articulation both trace to Mark H. Moore, whose Creating Public Value: Strategic Management in Government (1995) argues that public managers need what he called a restless, value-seeking imagination: an active search for opportunities to create public value, rather than a passive administration of an existing mandate. Moore's public entrepreneur differs from a private one in one decisive respect: a private entrepreneur answers, ultimately, to whether customers will pay and investors will fund; a public entrepreneur answers to whether an authorizing environment will grant and sustain permission, whether the value proposition is genuinely worth the collective resources it costs, and whether the organization can actually deliver it. In other words, an entrepreneurial public manager is not exempt from the strategic triangle — legitimacy and support, the value proposition, and operational capacity — but must satisfy all three at once, in real time, often while the idea itself is still being shaped (see Topic 1.2 — The Strategic Triangle: Legitimacy, Value, and Capacity). A public entrepreneur who wins capacity and builds a strong value case but never secures authorization is not innovating; they are freelancing with public authority, and the distinction matters however good the idea turns out to be.

Intrapreneurship is the closest private-sector cousin: entrepreneurial behavior by an individual working inside an existing large organization rather than founding a new one (see intrapreneurship). It is a useful comparison precisely because it shows what does not transfer: a corporate intrapreneur is usually judged, in the end, by a business case that a market will validate, where a public entrepreneur's case can only be validated politically and socially, through the authorizing environment described above.

Social entrepreneurship is the adjacent social-sector concept: individuals or organizations that develop, fund, and implement solutions to social, cultural, or environmental problems, often blending a commercial or earned-income model with an explicitly social mission (see social entrepreneurship). Much genuine public-sector innovation originates in the social sector — a nonprofit pilot, tested at small scale with philanthropic funding and less exposure to political risk than a government department would tolerate, that a public body later adopts, funds at scale, or absorbs into statutory provision.

Diffusion of innovations is the theory, developed by the American communication scholar Everett Rogers and first published in 1962, that explains how, why, and at what rate a new idea or practice spreads through a social system. Rogers described the population adopting an innovation as sorting, over time, into a recognizable sequence — innovators, early adopters, an early majority, a late majority, and laggards — a pattern usually visualized as the technology adoption life cycle's bell curve, and argued that adoption spreads through specific communication channels within a social system, not simply by an idea's intrinsic merit. Rogers's own later work extended the theory explicitly to policy: how institutional innovations are adopted, or not, by other institutions at local, state, or national level. This topic treats jurisdictional diffusion as this theory applied to public organizations specifically — the question of what makes a successful pilot in one city, agency, or country actually travel to another, as distinct from merely becoming well known there.

Diffusion across jurisdictions is not the same problem as diffusion across a market. A retailer that succeeds in one country can often expand into another by replicating its model with local adjustments, because the underlying legal and commercial environment — company law, competition rules, currency — is broadly comparable and the retailer itself controls the expansion. A public-sector pilot does not control its own expansion: the adopting jurisdiction must have (or acquire) the statutory authority to do the same thing, must build or already hold its own political support for it, distinct from the support the originating jurisdiction earned, and must have comparable operational capacity — staff, systems, funding — to run it, not merely a description of how the originator ran it (see Topic 3.9 — Intergovernmental Relations and Federalism for how legal authority is actually divided between levels of government, which this topic assumes rather than re-derives). A pilot that looks like a design problem when it fails to spread is often, on inspection, an authorizing-environment or legal-authority problem wearing a design problem's clothes.

Living labs are a related but distinct pattern: open, user-centered innovation environments where researchers, businesses, government bodies, and citizens co-create and test solutions in real-world settings rather than in a controlled laboratory (see living lab). Where a public-sector innovation lab is usually a team inside an institution, a living lab is usually a shared physical or virtual space — a neighborhood, a hospital ward, an online platform — in which multiple institutions and citizens co-produce and test an intervention together, a pattern with an obvious kinship to co-production more broadly (see Topic 1.5 — Stakeholders, Citizens, and Co-Production).

Design thinking is the set of methods — problem reframing, user research, rapid prototyping, iterative testing — that most public-sector innovation labs use as their working method, borrowed from and adapted from commercial product design (see design thinking). It is a method, not a strategy: design thinking can produce an elegant, well-tested prototype that still fails the strategic triangle if nobody has secured its legitimacy or checked whether the organization can sustain it past the pilot.

Best practices

1. Give every lab or pilot a mandate with an expiry date, not permanent exemption from evaluation. A protected space for experimentation is a resource the rest of the organization is implicitly subsidizing, and it should be renewed on evidence, not assumed to continue by default.
2. Insulate delivery risk, not accountability. A lab can be exempted from some procurement timelines or sign-off layers to move faster; it cannot be exempted from the same legitimacy and value tests every other initiative faces, or it stops being protected innovation and becomes an ungoverned exception.
3. Fund early pilots to learn, not to prove a foregone conclusion. A pilot designed so that only a positive result is reportable is not a test; state in advance what result would lead you to stop.
4. Document the assumed legal authority, political support, and capacity a pilot depends on while it is running, not after someone else asks to adopt it. These conditions are usually invisible to the people who built the pilot precisely because they already had them; write them down before they become someone else's missing prerequisite.
5. Judge a public entrepreneur's proposal against all three legs of the strategic triangle before backing it, not against the value case alone. An exciting idea with a strong value proposition still fails if nobody has tested whether the authorizing environment will actually grant it room, or whether the organization can deliver it past a small pilot (see Topic 1.2 — The Strategic Triangle: Legitimacy, Value, and Capacity).
6. Protect middle managers who take lawful entrepreneurial risks, including when the specific bet does not pay off. Distinguish clearly, in how failure is discussed afterwards, between a well-reasoned pilot that did not work and a genuine lapse of judgment or process; punishing both the same way teaches everyone to stop proposing.
7. Build formal peer networks across jurisdictions, not only published case studies. Much of what makes an innovation transferable is tacit — how a skeptical council committee was actually persuaded, which procurement clause was the real obstacle — and that knowledge moves through direct practitioner contact far better than through a report.
8. Before adopting an innovation from elsewhere, check your own legal authority and political support before checking your budget. Money is often the easiest of the three conditions to solve; a missing statutory power or a hostile authorizing environment can be the one that actually blocks adoption, and finding that out early is cheaper than finding it out after committing.
9. Publish negative and inconclusive results with the same visibility as successes. A sector that only hears about the pilots that worked systematically overestimates how often this kind of experimentation succeeds, and under-invests in the honest evaluation that would improve the average.
10. Separate the question "was the idea sound" from "was it implemented well" when a pilot disappoints. Many promising ideas are abandoned because of an implementation failure that says nothing about the underlying value proposition, and many weak ideas survive because a skilled team implemented them competently regardless.
11. Treat a successful pilot's route to permanence as a distinct design problem from the pilot itself. A team that is good at inventing is not automatically good at institutionalizing, and the handover from an insulated lab to a mainstream service line deserves its own explicit plan, owner, and budget line.

Questions to discuss with your team

Where in our organization does entrepreneurial behavior currently get rewarded, tolerated, or quietly punished? Most organizations have an informal answer to this question long before they have a written one, visible in who gets promoted after a bold idea fails and who gets sidelined. Use this discussion to surface real recent examples rather than abstractions — a specific manager, a specific proposal, a specific outcome. Consider whether the pattern you find matches the pattern your leadership would say it wants. If entrepreneurial behavior is punished in practice while praised in policy, that gap is worth naming before it is worth trying to fix.

Which of our current pilots would actually satisfy legitimacy and support if we tried to scale them tomorrow, and which are running on borrowed patience? A pilot can survive on a single sponsor's goodwill for a surprisingly long time without ever having earned broader authorization. Use this question to go through your live pilots one at a time and ask who, specifically, would need to say yes for each one to become permanent, and whether that person or body has actually been asked yet. Distinguish a pilot that is quietly building real support from one that is simply not yet visible enough to attract opposition. Discuss what you would need to do differently for each before a scale-up decision is forced by a budget cycle or a change of leadership.

If another jurisdiction or organization wanted to adopt our best current innovation, what would they need that only we currently have? This question inverts the usual diffusion conversation, which focuses on persuading others rather than on what the pilot actually depends on locally. Use it to list the specific legal powers, political conditions, funding sources, and staff skills your pilot assumes, separating what is genuinely portable from what is bespoke to your situation. Discuss whether you have ever documented these dependencies anywhere another team could find them. This is also a useful check on your own understanding of why the pilot worked, since teams often discover mid-discussion that they cannot fully explain their own success.

Have we ever adopted an innovation from elsewhere without first checking whether our legal authority, political conditions, and delivery capacity actually match the place it came from? Copying the visible form of a successful pilot — its branding, its workflow diagram, its staffing model — is much easier than replicating the conditions that made it work, and organizations under pressure to "do what worked there" often skip the check. Use this discussion to recall a specific past adoption, successful or not, and trace whether the underlying conditions were actually verified beforehand. Discuss what a proper pre-adoption check would look like for your organization specifically, and who would own doing it.

What happens to an innovation lab's ideas after the lab's own mandate, funding, or champion moves on? Labs are frequently built around a specific budget cycle, a specific sponsoring leader, or a specific piece of external funding, and their institutional memory can leave with any one of those. Use this question to trace what actually happened to the outputs of your organization's last lab or pilot program that ended, successfully or not. Discuss whether your current lab has an explicit plan for what survives it, or whether continuity depends entirely on whoever happens to still be in post when the mandate lapses.

Who plays the public entrepreneur role in our organization today, and do they have real cover to fail? Every organization that has produced a genuine innovation can usually name the specific person or small team who drove it, whether or not that role appears on an organization chart. Use this discussion to name that person or team honestly, and to ask what would happen to their standing if their next bet did not pay off. Consider whether entrepreneurial capacity in your organization currently depends on one or two individuals' personal risk tolerance rather than on any structural protection. Discuss what it would take to make that capacity survive the departure of whoever currently holds it.

In practice: a public value example

The transport authority of a mid-sized metropolitan region — fictional, illustrative of a pattern rather than an account of any real body — served a dense urban core well with fixed bus routes but had struggled for years with a scattering of lower-density suburbs where no route could attract enough riders to justify its cost. A junior service planner, frustrated by repeated proposals to simply cut the underperforming suburban routes, proposed instead replacing them with an on-demand minibus service: riders would request a pickup through a phone line or app, and a small fleet would be dynamically routed rather than following a fixed timetable.

The authority's leadership did not fund this as a full service; they created a small innovation team, gave the planner formal leave to run it, and authorized a six-month pilot in a single suburb, funded from a discretionary innovation budget rather than the main operating budget, with an explicit condition that the pilot would be evaluated against ridership, cost per trip, and rider satisfaction, and would end automatically unless renewed on the evidence. The pilot cleared the value case within four months — ridership in the pilot suburb tripled against the old fixed route's numbers, at a comparable cost per trip — but it also depended on things the planner had not fully appreciated while building it: a decade-old regional statute that gave the authority unusually broad discretion to alter service patterns without a full public consultation for anything under a defined budget threshold, and a genuinely enthusiastic city councillor whose ward included the pilot suburb and who had personally championed it to skeptical colleagues.

A neighbouring transport authority, in an adjoining jurisdiction with its own governing board, asked to adopt the model after reading the pilot's published evaluation. The technology and the operating model transferred easily. What did not transfer was the legal shortcut: the neighbouring authority operated under a different regional statute that required a full public consultation and board vote for any service change above a much lower threshold, adding nine months before a pilot could even begin. Nor did the political sponsorship transfer — the neighbouring board's chair was openly skeptical of a mode she associated, from an earlier failed private ride-share subsidy in her jurisdiction, with an unreliable service that would ultimately be judged a failure and used against her at the next election. The second authority eventually did adopt a version of the service, roughly eighteen months later, but only after building its own political coalition from scratch and negotiating a narrower pilot scope that its statute allowed without the full consultation requirement — a slower, locally adapted path that the first authority's success had made possible but had not made automatic.

Four sector lenses

Local government

Local government is usually where public-sector innovation is easiest to start and hardest to sustain: proximity to residents makes a problem and a candidate solution both highly visible, and a single council or department head can often authorize a small pilot without the layers of sign-off a national ministry would require. The constraint is capacity rather than authority — a small local authority may have the legal room to innovate but lack a dedicated team, a discretionary budget, or staff time to run a proper pilot, which is one reason many local innovations originate as informally as the fictional transport pilot above, championed by one motivated employee rather than commissioned top-down.

National government

National governments typically have the deepest resourcing for formal innovation labs — arm's-length digital units, policy labs, and innovation funds with dedicated staff — but face the hardest diffusion problem internally, because a pilot that succeeds in one region or one delivery arm must still cross the same legal, political, and capacity gaps to reach every other region or arm, now multiplied across a much larger and more heterogeneous authorizing environment (see Topic 3.9 — Intergovernmental Relations and Federalism). A national lab's central risk is producing well-evaluated pilots that never leave the capital, because nobody owned the harder work of adapting them to conditions elsewhere in the country.

Social sector and nonprofit

The social sector is frequently the true origin point of public-sector innovation, able to test a model at small scale with philanthropic or earned-income funding and far less political exposure than a government department would tolerate for the same experiment, and a government agency later absorbs, funds at scale, or writes into statute what the nonprofit sector proved first (see Topic 3.8 — Social Sector and Nonprofit Management for the funding models that make this possible). The characteristic failure here is the opposite of the public sector's: a nonprofit pilot can be too dependent on a single funder's priorities and a founder's personal energy to survive the transition to a government-run, statutorily grounded service, even after the underlying idea is thoroughly proven.

Multilateral and international

Multilateral and international bodies cannot mandate diffusion at all — they have no authority to compel a sovereign government to adopt anything — so their role is almost entirely knowledge brokerage: collecting, comparing, and publicizing innovations from many jurisdictions so that a government considering one has evidence and peer contacts rather than only a persuasive pitch. The Organisation for Economic Co-operation and Development (OECD)'s Observatory of Public Sector Innovation performs exactly this role, tracking innovations across member and partner countries and building the peer networks that let a practitioner in one government learn directly from a counterpart in another, without either government being able to compel the other to adopt anything.

Common failure modes

Pilotitis. An organization runs a long sequence of pilots, each individually well designed, none of which is ever formally scaled, ended, or evaluated against a decision to do either. The cost is a reputation for constant activity that never becomes lasting capability, and a workforce that learns not to take a new pilot seriously because the last several went nowhere.

Insulation as theater. A lab or unit is created with a public launch and a distinctive brand but no real mandate, budget, or protection from the organization's ordinary sign-off processes, so its "innovations" are, in practice, no faster or freer than anything else. The cost is cynicism, both inside the organization and among outside observers who took the announcement at face value.

Copy-paste diffusion. An adopting organization replicates the visible form of a successful pilot — its workflow, its branding, its staffing model — without checking whether the legal authority, political support, and operational capacity that made it work in the original jurisdiction are actually present in the new one. The cost is a launch that looks identical to the original and then fails for reasons the adopting team never diagnosed, because they were looking at the design rather than the conditions underneath it.

The unprotected entrepreneur. An individual or small team takes a genuine, well-reasoned risk on the organization's behalf, and when the specific bet does not pay off, they are penalized in the same way as someone who made a genuine error of judgment or process. The cost compounds over time: every future potential public entrepreneur in that organization has just watched what happens, and stops proposing.

Innovation detached from the value case. A pilot is pursued because the method is fashionable — a new technology, a widely publicized approach from another sector — rather than because a clear public value proposition was tested first. The cost is a well-executed pilot that clears every process gate and still fails to produce anything citizens would judge worth what it cost, discrediting the next, better-justified pilot by association.

These failure modes are not a special risk of innovation; they are the strategic triangle's usual failure modes, arriving in a newer costume (see Topic 1.2 — The Strategic Triangle: Legitimacy, Value, and Capacity). That is fitting for the book's last numbered topic: public value is never finished being created, only created again, by the next public entrepreneur willing to notice what is missing.

Maturity model

Checklist

• Confirm every active lab or pilot has a documented mandate, budget source, and renewal or sunset date.
• Confirm insulation from routine process does not also mean insulation from legitimacy and accountability.
• State in advance what result would lead a pilot to be stopped, not only what result would let it continue.
• Document the legal authority, political support, and operational capacity each live pilot currently depends on.
• Test every entrepreneurial proposal against all three legs of the strategic triangle before committing significant resource.
• Identify who in the organization currently plays the public entrepreneur role, and whether they have real cover to fail.
• Establish or join a practitioner peer network for sharing tacit knowledge, not only published case studies.
• Before adopting an innovation from elsewhere, verify legal authority and political support before verifying budget.
• Publish negative or inconclusive pilot results with the same visibility given to successes.
• Assign an explicit owner and plan for institutionalizing any pilot that clears its evaluation, separate from the team that built it.

Key sources

• Mark H. Moore, Creating Public Value: Strategic Management in Government, Harvard University Press, 1995 — the source of the public entrepreneur concept and the strategic triangle this topic applies it to.
• Mark H. Moore and Sanjeev Khagram, "On Creating Public Value: What Business Might Learn from Government about Strategic Management", Corporate Social Responsibility Initiative Working Paper No. 3, Harvard Kennedy School, 2004.
• Everett M. Rogers, Diffusion of Innovations, Free Press, first published 1962, later editions extending the theory to policy diffusion.
• Organisation for Economic Co-operation and Development, Observatory of Public Sector Innovation (OPSI) — https://oecd-opsi.org/.
• Nesta, "Public and social innovation labs" and the Nesta/Bloomberg Philanthropies i-teams research on innovation teams in government.

References

1. Innovation — Wikipedia — https://en.wikipedia.org/wiki/Innovation
2. Nesta (charity) — Wikipedia — https://en.wikipedia.org/wiki/Nesta_(charity)
3. Mark H. Moore — Wikipedia — https://en.wikipedia.org/wiki/Mark_H._Moore
4. Intrapreneurship — Wikipedia — https://en.wikipedia.org/wiki/Intrapreneurship
5. Social entrepreneurship — Wikipedia — https://en.wikipedia.org/wiki/Social_entrepreneurship
6. Diffusion of innovations — Wikipedia — https://en.wikipedia.org/wiki/Diffusion_of_innovations
7. Everett Rogers — Wikipedia — https://en.wikipedia.org/wiki/Everett_Rogers
8. Technology adoption life cycle — Wikipedia — https://en.wikipedia.org/wiki/Technology_adoption_life_cycle
9. Living lab — Wikipedia — https://en.wikipedia.org/wiki/Living_lab
10. Design thinking — Wikipedia — https://en.wikipedia.org/wiki/Design_thinking
11. OECD — Wikipedia — https://en.wikipedia.org/wiki/OECD
12. Mark H. Moore, Creating Public Value: Strategic Management in Government — Harvard University Press — https://www.hup.harvard.edu/catalog.php?isbn=9780674175587
13. Mark H. Moore and Sanjeev Khagram, "On Creating Public Value: What Business Might Learn from Government about Strategic Management" — Harvard Kennedy School Corporate Social Responsibility Initiative — https://www.hks.harvard.edu/sites/default/files/centers/mrcbg/programs/cri/files/workingpaper_3_moore_khagram.pdf
14. Observatory of Public Sector Innovation (OPSI) — OECD — https://oecd-opsi.org/
15. Public and social innovation labs — Nesta — https://www.nesta.org.uk/feature/innovation-methods/public-and-social-labs/