New executive agencies seem to be the government’s solution to too many quangos. Kate Jenkins argues that there are hard lessons from past experience with agencies which the government ignores at its peril.

The government should move with caution in setting up new agencies, (for example . the Department for Education’s Education Funding Agency and the Home Office’s National Crime Agenc) . A vast amount of public money flows through agencies in capital expenditure, transfer payments and running costs, and they employ about two thirds of the civil service. Agencies can make or mar the government’s relations with the public, meaning that Governments should manage them with serious care.

The first executive agencies were set up in 1988, to address Whitehall’s inability to deal competently with policy and management. The agencies (which eventually ranged from those with a small specialist staff, such as that of the National Measurement Office, to a huge nationwide network, such as Jobcentre Plus) were to be given a formal agreement with clear tasks, defined budgets and delegated powers. The senior management of the agency was charged with devising a system which would achieve the best outcome. They had to follow civil service principles and the objectives set in their agreement but not, as in the past, detailed civil service regulations.

The core of the relationship between department and agency was a ‘framework document’. This was, and should still be in some form, a statement of what the department wants the agency to do, how it will be organised, and the system for forward planning, budgeting, performance management and outcomes.

The department or ministry was to ‘interpret policy’ in a way that could be implemented, the agency was responsible for managing the implementation. This seemed fairly straightforward but relatively few people understood how difficult ‘interpreting policy’ could be. It took hard thinking and often uncomfortable compromises, for both ministers and civil servants, to take policy ideas and devise practical objectives and outcomes than could be implemented successfully.

When this design was followed, the agencies worked well. About 140 agencies were set up in the first seven years, ranging from the huge Social Security,  Employment Services and MOD agencies with budgets running into the billions to smaller specialised agencies like the Met Office and the QE11 Conference centre. Some agencies were amalgamated, some abolished and some new ones established. In 2011 there are about 80 agencies.

And there have been disasters – not many, but embarrassing, public failures which can be traced to dumping unresolved and difficult policy issues on an agency which was ill equipped or badly managed.:

• The Child Support Agency failed in 1994 because its policy was not completed and its staffing and its systems were not ready for use. The mess caused many parents and children hardship and misery.
• In 1999 the Passport Agency tried to rush policy changes, computer systems, the wrong time of year and inadequate staffing and created mayhem – and then passed the cost to its customers. The cost of a standard passport rose from £21 to £28in a single year (an increase of a third).
• The Rural Payments agency managed to fail on every count in 2006 – policy, implementation, systems and management – and passed much of the burden of cost to the people it was supposed to support. Its failure cost the farmers about £20 million.

An agency’s task was to implement, not define, policy. Definition was for the department responsible for the policy. These agency failures demonstrated the failure of its sponsoring department to carry out its primary task – to interpret policy so that it could be implemented effectively, and to check that effective implementation was done. The vast majority of agencies and their departments have done this successfully: many of the original agencies are still going strong.

The lesson for now is obvious. The agency will implement- not make good a failure to define what is wanted. It will find it difficult to deal with poor performance measures – one agency was given 900 activity targets and no outcomes by an over active but thoughtless department.

An agency is not a substitute for a weak department, it will not protect ministers from the consequences of bad decision taking, inadequate funding or poorly defined outcomes. What it should do, with the right framework, is provide a better focused service for its users and customers – which is what the public service should be there for.

Please read our comments policy before posting.

Kate Jenkins has written a report on the lessons from the past twenty years of agencies “Unfinished Business: the future for executive agencies”, for the Institute for Government”, published in July 2011.