Mulgan: setting out agenda for change

The director of the prime minister's Strategy Unit, Geoff Mulgan, explores the scope and pace of future Whitehall reforms.

All large organisations - whether governments, city administrations, multinational companies or armies - face two common problems. One is a problem of coordination: how to cajole and encourage an often huge flotilla of agencies, departments, units and professions to point in broadly the same broad direction, and at the very least not to undermine each other's work.

The other is a problem of organisation and integration: how to align incentives, cultures and structures of authority to fit critical tasks that cut across organisational boundaries.We coined the phrase 'joined up government' to refer to both sets of issues. Although some aspects of it are new - particularly the impact of the Internet - in other respects these are very old issues.

They faced all the big imperial bureaucracies whether Roman, Ottoman or Chinese and every military command attempting to coordinate complex forces.

In Britain similar problems led to the creation of multi-functional local government in the late 19th century as a joined up alternative to the separate boards for sewage, water, gas and education. In France (echoing ancient China) they prompted the creation of ENA as a tool for forging a 'joined-up' administrative elite.

Joined up government appears radical now in part because British government opted so firmly for departmentalism during and after its great expansion in the late 19th century.

A functional division of labour, with large vertically organised divisions or departments, held together by a relatively small head office, made sense not only for governments but also for large firms and city administrations in an era when communication and the management of knowledge were costly, and best organised within institutions and professions.

So government was divided into functions. Separate departments dealt with finance, education, defence, housing, colonies, trade and transport.

Often departments developed close relationships with particular professions: health with the doctors, education with teachers, the Home Office with the police, in line with Haldane's belief that the knowledge base was the best determinant of how organisational boundaries should be defined.

Funds were then voted by Parliament for specific ends, with tight monitoring to ensure that they were spent correctly.

Over time, however, the weaknesses of this model have become more apparent. The 'tubes' or 'silos' down which money flows from government to people and localities have come to be seen as part of the reason why government is bad at solving problems.

Many issues have fitted imperfectly if at all into departmental slots. Vertical organisation by its nature skews government efforts away from certain activities, such as prevention - since the benefits of preventive action often come to another department.

Many reformers in the past have tried to grapple with these problems. Prime Minister Edward Heath created a number of super-ministries. The Wilson government created the Joint Approach on Social Policy (JASP).

Many governments have set up cross-departmental committees like the one on inner cities under Margaret Thatcher in the late 1980s. Under John Major the Citizens Charter initiative attempted to introduce common principles across all service delivery departments.

At a more local level innovations included the Single Regeneration Budget.

But these reforms had only limited impact. As a rule government works best when there are clearly identified critical tasks; when authority and resources are distributed in ways that enable these to be carried out; when there is a clear sense of mission from top to bottom; and when there is sufficient freedom and flexibility for those working as managers or front-line delivery to get the job done.

Many of the previous government reforms had not met these conditions.

That is why more recent reforms have concentrated on reshaping all the factors that drive behaviour in government: reforming the way money was allocated - to ensure that more of it was allocated to specific problems, areas, or client groups rather than to functional bureaucracies (as has now been done with budgets for children, drugs, neighbourhoods and conflict prevention); the way career rewards were organised - rewarding those who acted corporately or collaboratively with promotions, honours and bonuses; designing targets that would be shared across agencies (a large proportion of PSA targets are now shared); new units like the SEU; new local structures like Youth Offender Teams; reshaping services around customer needs (as UK online has done); tackling the day to day cultures of the professions; ensuring that information and knowledge was shared better at all levels; giving politicians kudos for joined up roles as well as traditional vertical ones.

There were, and are compelling reasons for driving forward with reform. One is that many of the problems that most concern government - competitiveness or crime - are inherently ill-suited to existing departmental structures.

Others include the growing evidence of the interconnectedness of problems- for example the links between poor school performance and family factors; technology which now makes horizontal communication far easier than in the past; consumers wanting services fitted to their needs rather than administrative convenience.

What will the future bring? Are we at the early stages of a fundamental transformation of government, or will joined-up government turn out to be just another fad? My own guess is that although governments are necessarily quite conservative institutions the pace of change is unlikely to let up. It is unlikely that government will ever be predominantly organised in horizontal as opposed to vertical structures. If it was there would be as many boundary problems are there are today.

Instead the future shape of government is likely to involve a combination of vertical hierarchies, particularly for carrying out long-standing tasks with clear lines of management and accountability, and horizontal structures for determining strategy and carrying out shorter-term tasks. As well as determining strategy and overseeing performance the role of the centre of government will continue to be that of allocating the key resources at its disposal - money, people, political capital, legislative time, knowledge - to both vertical and horizontal parts of the system.

In effect that would mean government evolving further in the direction it is already taking. It would involve:

• More work becoming project based.

• More policy making being done in a cross-cutting way, but with the close involvement of practitioners.

• More of the budget being tied to outcomes - and then allocated across departments and agencies according to how much they can contribute to outcomes.

• More vertical functions being passed out to agencies, leaving behind slimmer, but more integrated central staffs.

• A much greater emphasis on shared knowledge management as the glue holding central government together.

• An expectation that civil service careers will move across and beyond government.

• Use of the integrative power of the Internet to organise access to services according to people's needs rather than producer convenience.

• A much more energetic approach to reshaping business processes that cut across departmental boundaries.

• A steadily growing role for local partnerships in integrating the work of both national and local agencies on the ground.

• A greater emphasis on professional formation across boundaries.

Longer term more radical options may also be feasible. Some have advocated that responsibility for whole systems - like the criminal justice system, transport or children's services could be organised in an integrated way, potentially with purchaser-provider splits, rather than, as at present, divided between many different agencies and professions each with their own budgets, structures and targets.

My view is that it is right to continue with an evolutionary approach rather than a big bang. However, over time the biggest gains will come from moving beyond the relatively modest joining up of the late 1990s to more fundamental systems redesign.

Joining up in all its forms has happened, is happening and will happen even more in the future. It may rarely if ever be perfect. But governments that can think and operate in 360 degrees will over time prove better at solving problems and meeting needs than governments that remain trapped in the vertical hierarchies that they have inherited.

This article is an edited version of a talk given to the British Academy.