An essay on the future of government services by Tom Bentley

Essays extolling the need for governments to get ‘connected’, lateral, vertical and all that kind of stuff – the need to find new models to engage stakeholders and to break down the silos of departments – are not usually my cup of tea. My problem with them is that as commonsensical as these ideas sound, they are frequently pretty superficial. And they don’t show much awareness of the reasons for why one would arrange things in silos in the first place. (Those reasons would include trying to create manageable lines of accountability.) Budgeting is the same kind of exercise – an exercise in self constraint which doesn’t make sense until you understand it in the context of building lines of accountability in a boundedly rational and knowledgeable world.

In any event I was impressed when I read this conclusion to a volume recently published by Demos by Tom Bentley. It’s not that it proposes anything new, but I think the way it discusses the issues is mature and insightful. I’ve also become more and more convinced of the need for governments to develop much greater levels of responsiveness to their evolving environment.

The world is increasingly complex and it seems to me that there are two responses to this. A right of centre ‘free market’ approach takes the view that governments growing increasingly complex is an inherently bad thing. Governments should stick to their knitting – which is providing frameworks of law and basic public goods like defence. The problems with this approach are firstly, that the world is becoming more complex and so governments being embedded within our society and economy are drawn into that complexity. For instance if they want to engage in large financial transactions which they pretty much have to, they will need to master their complexity. Secondly and likewise, as the market becomes more complex it becomes more complex to regulate it. The corresponding kind of approach to regulation involved in this mind set is ‘minimum effective regulation’ in which governments do the least they need to achieve their objectives. The problem is that commonsensical as it sounds, no government in the Western world has managed to tame the inexorably growing complexity of regulation with the policy of minimum effective regulation.

The alternative – which might loosely be regarded as left of centre – is the pathway set out here where governments make themselves more responsive to their environment and the various stakeholders. Where they try to develop broader, looser ways of collaborating. The problem is always how to do so in a manner that improves regulation and service delivery (rather than makes it worse – not necessarily an easy thing) And one must do so further constrained by the need to do so consistently with being accountable to the public interest rather than private interests.

Anyway, I enjoyed the essay and asked Tom if I could post it here, which he agreed to and – which I’ve done over the fold.

Evolving the future

Tom Bentley

Late in 2006, the Australian Federal Government published an advertisement in the Australian, the country’s national newspaper. It announced a new agreement on social security between the Kingdom of Norway and the Government of Australia, beginning on new year’s day 2007. The agreement begins, “Wishing to strengthen the existing friendly relations between the two countries, and resolved to co-ordinate their social security systems and to eliminate double coverage for workers, the parties have agreed…”

If I meet pension requirements in Norway, I can still receive the benefit if I then move to Australia. If I have worked in both countries, I can add together the years and put them towards entitlements in whichever country I retire in. While the systems remain separate, it becomes possible to personalise my participation in them.

This kind of activity is growing. It illustrates how greater interconnectedness across the world prompts collaboration between governments – separate sovereign entities – in order to solve shared problems and make life easier and better for citizens. Social insurance and welfare, for so long conceived as the product of different nation state systems, are becoming internationalized. The governments do not merge their schemes, or try to run them in exactly the same way. But they agree that time spent working and paying taxes in one country can be treated as equivalent to doing the same in another.

As bilateral agreements between specific governments add up, they slowly contribute to the formation of an institutional environment – a space shaped by rules – which can be far bigger and more significant than any one of its formal elements.

These are collaborative relationships between governments, but the same principle applies to collaboration within societies, and within governments themselves; the state is not a single, monolithic entity, but a proliferation of organizations, teams, interests and centres of power. Collaboration between government and society, and between states seeking peace, wealth and security, is as old as states themselves. It is part of the process through which governments have emerged, but it is becoming more important for three linked reasons.

The first is the growth of connectedness, or connexity. If every problem is connected to something or someone else, then collaboration to solve it is logically necessary. Second, networks, especially the internet, make collaboration easier, cheaper and therefore more diverse and wide-ranging through a range of tools, practices and cultures.

Third, reform of the state over the last thirty years has, as Sue Goss points out, pluralized and multiplied the number of agencies involved in public service provision. Privatisation, contestability and decentralization mean that, where government is seeking to create a public good, it is increasingly likely to do it through collaboration with organizations in other sectors. Changing citizen expectations; of less deference, more flexibility and better service, reinforce this shift.

Vertical and lateral

We as citizens have become used to an image of government which is separate from the rest of society, defined by its coercive nature and its martial roots, logically distinct from the worlds of market and civil society. There are good reasons why governments should be perceived as such, and why they should want to be – to achieve the impartial administration of justice, for example, and to maintain the monopoly on the legitimate use of force by avoiding the capture of power by specific interests in society.

In fact, the need to collaborate is designed into government as a result of democracy and constitutionalism. The separation of powers into distinct, independent entities is the ultimate political design principle for collaboration. It seeks to ensure that no one agency or clique can impose its own priorities wholesale. To put boundaries around institutional authority, we need defined functions and vertical powers. To create solutions across complex fields, they need lateral relationships and capabilities.

Combining these effectively leads to successful government. The consequence is that successful politics and policy require persuasion, bargaining, compromise, sharing of benefits and, even if indirectly, learning between different players and territories.

This need to combine specialization and integration, command and consent, competition and collaboration casts fresh light on the value of federal systems of government, and points to why they have emerged as a way to balance the competing interests and identities of separate communities with the interests and needs that they simultaneously share. As Robert Wright wryly notes:

“In 1500 BC, there were around 600,000 autonomous polities on the planet. Today, after many mergers and acquisitions, there are 193 autonomous polities. At this rate, the planet should have a single government any day now.”

But while the force of history encourages unification, the merger process has been accompanied by enormous growth in the lateral connections and relationships used to manage across and between governments, giving them adaptive flexibility alongside economies of scale. As a result, institutional design has an enormous impact on how a given system solves collective problems. Federal systems such as the US, Swiss Canadian and Australian, designed pragmatically to give a self-balancing weight to different constituencies, can encourage both competition and collaboration between members of the same federation with positive-sum consequences.

Renegotiating the terms of federation, in order to achieve structural reform which creates positive sum economic and social effects, provides us with a clear example of the benefits, and the difficulty, of achieving collaborative governance. Australia’s current National Reform Agenda, in which federal and state governments of different parties have committed to negotiating shared investment in reforms designed to boost the long run capabilities of the Australian population through human capital, regulatory and infrastructure investment, provides a working example.

Beyond the current options

All these reasons help to explain why governments have moved significantly towards a fresh emphasis on collaboration in the last decade – an emphasis that is moving from a policy focus on improving ‘service’ towards the issues of personalisation and coproduction that require more radical redesign of services and new organizational forms.

The organizational designs which government can draw on to pursue these relationships are also expanding in range, from contract management and Memoranda of Understanding to joint ventures and a range of network designs. Sir Michael Barber, pioneer of public service reform, recently argued that there are essentially only three models of reform; command and control, quasi-markets; and a ‘combination of devolution and transparency’ in which governments delegate to or contract with service providers and then hold them accountable.

But there is a much broader range of system models and reform options available if you recognize the range that can evolve, or emerge, from different combinations of Barber’s three basic types. If you build an architecture for collaboration, as well as competition and control, and recognize that the strategies of all organizations are likely to evolve in response to changing conditions, then a far more diverse range of possibilities comes in to view.

This broader view allows us to recognize the range of platforms that government can use to offer services, and the combinations of organizations that can be involved in them. Eggers, for example, identifies channel partnerships, information dissemination networks, supply chains, service contracts and ‘civic switchboards’ on their spectrum; the burgeoning science of networks could provide many more.

Yet even this range does not cover what is arguably the most important area for the future of public services: the role of government in shaping an environment through which citizens themselves can collaborate and produce various kinds of good. This, matters because the social and economic conditions that drive collaboration reinforce the need for governments to go beyond their current institutional options.

Citizens innovate through collaboration

These are the conditions that lead Yochai Benkler to advocate social, or ‘commons-based’ production as the most important new way to meet diverse human needs. The same set of broad changes leads Charles Leadbeater to emphasise the possibilities of mass creativity, or ‘We think’, in which many institutional and economic barriers to collaborative problem solving are broken down and collaboration for mutual gain can happen on mass scale and at great distance in everything from the organization of work and the production of energy to the provision of education.

Benkler argues that this shift allows many more ways for people to meet their own needs by creating services, activities, culture for themselves. He also maintains that these production processes inevitably draw on the resources generated by the creativity of others. These resources are the ‘commons’ from which we find raw materials to shape our own personal efforts, as well as the comparisons and sources of inspiration which we use to ground our sense of who we are and what we want.

Social production is happening already; in informal networks of learning, social care and work coordination, in sports clubs and local health centres where shared social activities contribute to wellbeing and to better health outcomes. It is intertwined with the architecture of professional service delivery, and often obscured by the collection of statistics which highlight only the more formal processses.

The dominant institutional frameworks through which public responses to human need are pursued; the competitive field of the market and the command-based domain of the state, are too narrow for the reality created by an evolving society. That is why partnership and joint venture have become part of the government repertoire. But crucially, the new forms of production can evolve into larger scale structures capable of supporting mass-scale activities, and therefore competing with the scale of industrial production or of government procurement, but using quite different rules of participation. As Benkler puts it:

“These architectures and organisational models allow both independent creation that coexists and coheres in usable patterns and interdependent cooperative enterprises in the form of peer-production processes.”

Traditionally, the scope for activity driven by ‘non-instrumental’ motivations is ascribed to the civic realm and third sector of non-profit, non-governmental organizations (and to the private realm of family and friendship networks). But the boundaries of this category are fuzzy, ranging from the tiny to the multinational, and intersecting with state funding and market trading in numerous ways. The emergence of collaborative production means that such ‘social production’ processes can be intertwined with activities and institutions that are grounded firmly in both market and state, not hived off artificially into a catch-all category of third or community sector organisations.

The significance of this shift for the shape of what we currently call public services is huge, given the stage that public service reform has reached in many countries. The monolithic, catch-all provision of the past is widely recognized as an undesirable platform for the future. But the current range of reform options, particularly those focused solely on the privatization of services and assets and shifting the burden of risk onto individuals, is equally unpalatable.

Yet in countries which are ageing, diversifying, and shifting further towards service-based economies, new ways to intertwine productive economic capacity with social investment are urgently needed. The greatest need is exactly at the interface between self-provision by individuals and families and formal service provision;

This is partly because the pattern of human need is shifting, in wealthy industrialized societies, in ways which make the traditional methods of paying for public services unsustainable, and the traditional methods of organizing and regulating them ineffective. As the burden of disease shifts towards the chronic, and the nature of work becomes intertwined with the expectation of continuous learning, new patterns of production for these goods are needed which can simultaneously personalise and replenish the commons on which they draw. This should be the goal of the collaborative state.

A glance at some of the other pieces in this collection, and at Demos’ past catalogue, shows that this is exactly what is happening, from below, in the field of public service provision. Collaborative service design by organizations operating together in local areas is the foundation of effective co-production between citizens and government. The collaborative state has to include those organizations and networks that can mediate between individual need and universal rules; it is through that process of mediation that service can be personalized, responsibility shared and value co-created.

But two great historical barriers stand in the way of governments’ ability to practice and promote collaboration across all their functions. The first is that the establishment of modern, reliable, professionally run states rests on the ability to prevent corruption, which is a form of illicit collaboration. This means that many of the protocols, routines and instincts and routines of government are dedicated to screening out unwanted contact, or channeling it through dedicated routes, reinforcing what can be experienced from the outside as rigid and opaque.

Second, government grew through the 19th and 20th centuries around the idea of functionally based, professional services in which expert knowledge was organized into separate units – silos – and governed through vertical chains of hierarchy and accountability.

An unholy alliance of history, accountability and power combine to hold this approach in place. Of course, preserving accountability to parliament and encouraging responsible use of public budgets is important, especially in complex systems. But there is a simpler reason for the stasis, one which public servants and politicians can rarely own up to in public: the struggle for power. One former Conservative Cabinet Minister remarked to me soon after the 1997 election that trying to reform Whitehall departments meant dealing with ‘feudal baronies’, a remark that chimes uncomfortably with Henry Tam’s observation that the barons have become postmodern under New Labour.

While political power is measured by the size of the departmental portfolio, and civil service careers progress towards the pinnacle of hierarchy through control of ever-expanding chunks of organization, the tendency towards organizational co-production at the top of government is always going to be limited. This, of course, is well known, but how to overcome it is not. It matters not so much because everything depends on these tiers of government, but because they reinforce a culture and a set of assumptions which weaken the possibilities of collaboration elsewhere.

Officially, government still lives in a Newtonian world where every reaction produces its own absolute effects, which should be separable and measurable in isolation from all other activities. In this world, policy rationally sets the objectives of delivery organizations, allocates resources, management control and accountability, and the outcomes of, say, a hospital reorganization or a crime reduction target should be achieved through the vertical transfer of instructions and incentives down and up the chain of command.

This tendency is reinforced by the ‘principal/agent’ mindset of the New Public Management, in which the strategic task is always to establish who is really in charge, a precondition allocating ‘operational’ accountability. But as Charles Sabel argues, the separation of strategy from execution is repeatedly undermined by the realities of implementation and the fact that the operators are usually those with the greatest circumstantial knowledge about how things work.

The erosion of these assumptions is part of a much deeper shift in our understanding of the nature of organizations, away from the attempt to make them work like machines following commands, and towards a view of more complex sets of relationships, in which people act for a mix of motivations and where change arises from both conscious, formal decision making and from a constant process of adaptation, adjustment and improvisation.

Beyond this shift, as Sabel points out, “the canonical form of this organization is federated and open”. While higher level organizations (parliament, government departments) set general outcome goals and boundaries of action, the ability of the overall system to find effective solutions, and to adapt successfully to changes in the external environment, depends on the ongoing interaction between rule setting from above and lesson learning, in the light of experience, from below.

As Sabel argues, “These federated organizations respond to the problem of bounded rationality not primarily by decomposing complex tasks into simple ones, but rather by creating search networks that allow actors quickly to find others who can in effect teach them what to do because they are already solving a like problem.”

A collaborative state is one which can reshape its own actions, investments and architecture around this search for continuous improvement through learning.

Evolving the future

But can government really embrace such a future? Private firms are arguably far more comfortable in a Darwinian world, not least because survival of the fittest is an accepted principle. Can the art of governing develop into the capacity to design rules and project goals for complex sets of organizations, learning systematically from their efforts and designing regimes for collaboration that maximise the public value they create?

The range of current practice suggests that collaborative innovation is rich, varied and growing. The growing difficulty of maintaining traditional service models will continue to prompt innovation from below. Much harder to achieve, though, is the adaptation of large scale institutional architecture. But even here the future of collaboration is more likely to evolve from the growth and spread of new practices than from wholesale structural change imposed from above.

The key is to understand how to use policy design and the management of implementation to model, incentivise and then learn systematically from patterns of collaborative action. As these approaches become more visible and more successful, the feedback they create on what succeeds needs to be channeled systematically into the recurrent decision-making cycles such as budget allocation. Unfortunately, the connections between evidence, practice and budget allocation remain weak in most systems.

But opportunities to reshape the state through collaboration abound. There is no reason, for example, why American cities, Australian states and the EU should not collaborate to develop solutions to climate change through carbon trading; it is already beginning to happen. Equally, it should become a core part of government’s role to co-design and invest the architecture, enabling the wider public realm of institutions and organizations to collaborate in ways which make co-production, or social production, a visible feature of everyday life.

Governments can do this by:

– Redesigning public procurement processes to encourage federations and network-based consortia to come forward with innovative solutions to cross-cutting public needs

– Experimenting with changed departmental structures based more heavily on teams and projects, which reward effective cross-organisational collaboration and make senior managers accountable when it fails.

– Adjusting parliamentary accountability regimes to seek evidence of learning and intelligent explanation, rather than mechanically searching for proof of the gap between rule and reality.

– Building ‘open architecture’ designed to make collaboration easier by helping public agencies, firms, civic organizations and so on to find each other on the basis of working on similar problems.

– Investing in modelling and forecasting techniques which examine the behaviour of complex fields of agents adapting to various conditions and environmental changes, rather than the limiting assumptions of classical economic theory or the linear predictions of traditional implementation planning.

– Seeking to design public agencies capable of taking a long term, population based approach to the outcomes they seek – for example in preventive health care, and rewarding them with assets and new responsibilities in return for long term outcome improvement.

– Building ‘learning systems’ which seek to nurture and scale up innovation through rapid cycles of design, application and feedback across groups of organizations working on a common problem, and rewarding consortia that come up with successful innovations.

The collaborative state mixes up many roles, powers and assumptions that have held for more than a century of modern government. But the forces undermining these modern myths have already been unleashed. Reformers are already seeking new routes through which to achieve large scale change, and new models for collective provision in diverse societies. These new patterns, driven by both collaboration and competition, will emerge from below.

Policy, regulation, funding and learning systems then have huge impact on how they are taken up and spread, and who gets access to the value that they create. Collaboration, pursued with discipline, is the route to the redesign of our large scale services and governance structures. The challenge of leadership is to focus it on the problems that government exists to solve.

Tom Bentley is an Executive Director in Victoria’s Department of Premier and Cabinet and Director of Applied Learning at ANZSOG, the Australia and New Zealand School of Government. He writes here in a personal capacity.

This entry was posted in Economics and public policy, Politics - national. Bookmark the permalink.
Notify of

Newest Most Voted
Inline Feedbacks
View all comments
16 years ago

This is a very interesting article and appears to set out the next narrative after the ‘Steering and Rowing’ 90s mantra (but much more diffuse and complex).

I must confess that I remain rather cynical about the chances of success in much of this (a healthy cynicism being an essential tool of a public servant). The changes appear to rely on flexibility, risk and innovation, not something bureaucracy is known for.

What will be the drivers to implement this? It looks hard and risky. Are there some successful models where this has been done?

16 years ago

I’ve spent most of my working life in the “3rd sector” – in social services – and have a predictable view that favours a strong public sector.

However, as I read this thoughtful article I reflected on a couple of things that happened to me in the last couple of weeks. Both of them show to me that the people within government, and probably the culture within which they work, are ill-prepared to support innovation and collaboration in the future Tom describes.

I went to a (nameless) State government department pre-tender briefing for a largish community service project. The tender process they proposed was an Expression of Interest, 3 agencies short-listed, all to go away and prepared a detailed response to brief including the amount of capital they would bring to the project. The department would pick the proposal they liked best, or even meld two of the proposals and then put the proposal out to open tender. They genuinely could see no reason why we wouldn’t give our IP away for free. This is a environment where most community services now go out to competitive tender and the sector has been warned against collusion.

Keating brought competition into our sector and I’ve seen it be the enemy of collaboration. I’ve also seen government bureaucrats in social services unable to come to grips with governance in the not-for-profit sector. And that’s often because they have no experience of what good governance is. I agree with Jonno – I’m not sure the skills are there and, even though the macro policy is there, it’s not supported at the interface.

Tom Bentley
Tom Bentley
16 years ago

Both these responses reflect the realities of trying to operate with government as it is, and the difficulty of moving in the direction I outlined in the essay. I think it is right that often, governments in all parts of the world struggle to ensure that their assumptions about how to operate keep up with the methods they are using. I’ve repeatedly come across examples of innovators who are stopped because government cannot find a way of protecting or valuing the stake that they bring to the table. This partly reflects the inadequacy of our IP framework, and the fact that it is mostly designed to protect the interests of very different kinds of commercial innovators. But part of what I’m arguing for is that the overall collaborative direction can be pursued in an evolutionary way – creating change by creating examples and building them towards critical mass, rather than going straight for the chicken and expecting it to change its whole pattern of egg production as a precondition of any other change. It’s often frustrating for those on the outside, but often I think it’s most effective route to change.

16 years ago

Presumably a key easy to measure indicator of the extent to which this takes place will be the size of Government Departments.

The current status quo is super-departments and large bureaucracies are not noted for their flexibility. The key will be resolving the tension between central control (/ responsibility?) and allowing agency flexibility.

We will see.

paul frijters
paul frijters
16 years ago

This is not my main area, but I think its important enough to give Tom a decent stab at an answer even if it is based on rather vague knowledge.
As an economic student, you get fed on the theories of Niskanen and his ilk, in which bureaucracies are made up of fiefdoms warring over policy territory and forever trying to expand at the cost of other interests. Loyalty to the state and political allegiance holds it together at the centre a bit, but any room in any particular fiefdom is quickly pinched by someone else who will hold onto it tightly by being as inflexible as they can be. I suppose I still believe these stories and find it hard to think of the bureaucracy as a place where flexibility could ever flourish because to be flexible is to open up opportunities for having your budget cut and your policy territory invaded by someone else. Its administrative suicide to be flexible in what you do. To be more precise, your best strategy is to be flexible in terms of taking on more tasks for your empire, but inflexible in letting go of existing tasks. To some extent, being inflexible is the strength of a bureaucracy: you can count on a bureaucracy following the rules in almost any circumstance, which makes the rules you start out with believable in the eyes of the rest of the population. Putting it simply: Do you really want a police thats overly flexible?

The only way to have flexibility in terms of policy space would be to have unattached civil servants who are not interested in running fiefdoms and are quite happy to see their niche disappear when the societal need for their particular service disappears. Thats what real flexibility means: being without a departmental home and without any task you have some implicit territorial feeling about. That comes at a cost though: if you do away with the loyalty to the local fiefdom, you do away with the benefits of that loyalty, such as team spirit, institutional memory, long-term specific investments, and a sense of accountability to the beneficiaries of ones actions. It would then all be about appearances. What Tom in my interpretation seems to want to see is a new kind of human. Someone who feels no unhelpful loyalties to the department he or she happens to be in, but is nevertheless fully committed to the public cause and thus willing to take risks, be innovative, and change jobs all the time. Are we thinking about genetically engineering a new human to be like this, or do we think that people nowadays are better able than before to quickly shift local loyalties when ‘super’ loyalties dictate this? If not then how are we going to mould such a new human?

As to risk, the usual political science story seems to apply: political masters will take credit for what you as a civil servant do well and blame you if things go wrong, creating the incentives for you to be extremely risk averse. That basic asymmetric incentive structure cant be broken without taking politics out of it and giving property rights over an aspect of the bureaucracy to something else. That’s undemocratic though so getting rid of risk aversion is another no-goer without major upheavals. Innovation is not something a bureaucracy would necessarily oppose, but when innovation is inherently risky it of course does become opposed for the reason cited above.

Hence I find myself in the camp of the sceptics. I have trouble seeing how risk, flexibility, and innovation can be brought into the civil service without radical new governance models. The usual suggestion of the economists to these standard issues still applies though: outsource as much as possible because you can change the people you outsource to relatively quickly and its easier for the outside organisation to take risks and innovate since they can charge for the improvement in productivity. Youll never get that kind of flexibility within a bureaucracy because of imperfect property rights. And if the outsourcing itself runs into flexibility problems like Angharad suggests, then you may want to think about outsourcing the whole thing a level up by outsourcing the whole area of policy rather than a specific project. Rent a transport department for 20 years rather than just a transport firm for one particular bridge. Now, theres an idea.

Nicholas Gruen
16 years ago


Your analysis is all pretty black and white. Life isn’t like it is in the models – it really isn’t. Of the things you’ve written, I guess I agree most with the stuff about risk aversion. As for the inevitability of fiefdoms, well it’s a tendency I admit, but often bureaucrats don’t want fiefdoms. Many are genuinely not into empire building. They are broadly speaking trying to a good job – but certainly that’s as confined by ideas of due and proper processes being followed. Often doing a good job involves solving problems that are served up to them. Thus the endless expansion of regulation – as the problems mount up from the media the pollies etc.

If we redefine the way problems are served up – or encourage those more thoughtful bureaucrats to respond to the problems that are served up in a way that is more open and facilitative rather than prescriptive – we might make a bit of progress.

You’ll note in Tom’s comment that he’s suggesting something pretty incremental – not the development of a new breed of human.

16 years ago

Nick – a very nice response indeed.

Paul – we already outsource some policy development. This is where we get consultants to undertake reviews/develop strategies etc. You will find enormous numbers of KPMG, PwC etc projects on the lists of consultancies prepared by Departments. In fact, those firms are now a key interchange place for executive public servants.

Of course we retain control over the process (after all we may be unelected but we do report to the elected Minister). We also often happen to be the experts not only in some of the content but also in what will work in the existing system.

Furthermore any reading of history (from colonial or even Roman times) shows the continual remaking of the public service (the 90s were a particularly vivid example).

That all said, I would be rather interested in some actual examples of this new paradigm at work.


[…] Era Governance” agenda that Patrick Dunleavy has been promoting at the LSE as well as the work of Tom Bentley at Demos on the future of government services. The Report advocates the following […]