Knowing better when to start and stop

How to start and how to stop: two problems with trying out new things in the government. Both are important for taxpayers.

By the time an idea materialises into a “pilot”, so much may have been invested in research, planning and persuading stakeholders, other officials and ministers that it is almost predestined to evolve into a full-blown operation.

Rigorous evaluation could fix that. And departments do have evaluation programmes. The Ministry of Social Development (MSD), for example, has a centre for social research and evaluation, but evaluation appears to be a small part of its work. MSD’s website’s “introduction” to the centre is dated 2004 and lists seven programme evaluations. The latest bulletin of a committee in the centre charge with fostering research and evaluation (SPEaR) is dated 2006.

One difficulty is cash. Even in flush years cash is limited. In a crunch officials can more readily postpone assessment than curtail action.

Ferreting out ineffective programmes is part of what Bill English has been on about as he steers the budget back to balance. Chief executives were instructed to prioritise programmes for value, value-for-money and achievability and, logically, drop those with the lowest priority.

English and the Treasury have also pushed departments to mine suppliers and contracted service delivery agencies (and their own staff) for new ideas on how do the government’s work.

English and Treasury Secretary John Whitehead have told public servants that if they can’t devise more cost-efficient ways of doing their job and private firms or not-for-profits have better ideas, the outsiders should get the job. Departments have been told to join up more, in ensuring citizens’ easy access face-to-face or electronically, in delivering programmes and in buying in equipment and supplies. That favours bigger suppliers. Much of this is not new but the pressure is.

There are thousands of contractors, especially in social, health and education services. They range from having decades of operation and hundreds of staff to recent startups with a handful of amateurs. Managing this diversity eats up departmental resources. In opposition National urged rationalisation, though that risks locking out social entrepreneurs with bright, viable ideas.

Add in the complex contracts which departments have devised to limit the risks in managing outside agencies. A medium-size outfit — say with around 60 operatives — might have a dozen or more contracts, each for a specific activity. Managing those contracts consumes agency time that could be spent in the field.

Social Services Minister Paula Bennett wants a move to simpler outcomes-based contracts, under which an agency undertakes to produce a measurable result and is judged on the result.

A prototype is the three-month “pilot” contract with Auckland family violence agency Shine to work with police and Child, Youth and Family and help families in the less acute cases. It will be evaluated after the trial.

The word “prototype” is important. Christian Bason from Mindlab in Denmark, who ran workshops with state sector managers in December, argues that a “pilot” is “often the beginning of scale” and not a trial but a “solution”, with the risk already extracted. Bason says true reform requires risk: small-scale prototypes which are experiments and on which quick feedback is “harvested” by evaluation.

The “prototype” notion fits well with social entrepreneurship, championed by the Social Innovation Network which sponsored Bason’s visit.

In the economy, entrepreneurs succeed and fail. So do entrepreneurs in social services. But, since most draw at least some of their funding from the government, the possibility of failure prompts ministers and departments to be risk-averse in backing them. That risks blunting entrepreneurship.

A short trial, properly evaluated, might be a way to cut losses from a failure while avoiding stifling potential successes.

To get quality and consistent evaluation, Bennett has won some support from ministers for a proposal to set up a separate unit to bring together the evaluation and associated research done separately by departments. A loose parallel might be the Education Review Office (ERO) which audits schools’ educational performance and is independent of the rest of the education bureaucracy.

A separate agency would have a budget dedicated to evaluation, with no scope to divert funds. It could promote transparency and truncate experiments and existing programmes that don’t deliver specified outcomes. And by separating assessment from the original decision to start a programme, contract an agency or back a social entrepreneur, it might give departments more latitude and encourage more experimentation.

In short, starting and stopping things in the government might get modern.