This report argues that rather than outsourcing ever more public services, governments should introduce social impact bond models within the public sector.
In one of his first major steps as Prime Minister, Tony Abbott has delivered on his promise to establish a Commission of Audit into the Commonwealth public sector. The exercise is in the image of similar commissions called by incoming Premiers in Queensland, New South Wales and Victoria.
The Commission will find, as its state counterparts have done, that the public sector exhibits poor productivity and should be reduced in size, with important services to be outsourced to the private sector. That this finding can be predicted so confidently in advance should raise questions over its validity. In truth, public sector productivity is a nebulous concept and is being used as a smokescreen to cut spending by governments unwilling to rebuild a broken tax base.
How do you track outputs per hour or per dollar when the desired outputs are so hard to measure? How does one 'price' a well-educated child or a rapidly cured patient?
Since they are unable to do this, governments instead adopt narrow 'target' measures as proxies for productivity, such as hospital waiting times and standardised test scores. But a focus of these narrow measures often results in the target being achieved at the expense of other equally desirable policy outcomes.
Faced with this, governments resort to an 'efficiency dividend', an annual budget cut (typically 1-2% p.a.) which assumes public service agencies can deliver the same outcomes with less resources each successive year. In the absence of hard data on outcomes, the efficiency dividend acts as little more than a crude cost-cutting
What then is a sensible way forward, which recognises the need for accountability within the public sector while acknowledging the limitations of private sector productivity measures? This report argues that answers might lie in the not-for-profit sector, where a range of new approaches to performance are being developed.
Foremost among these is social return on investment, which seeks to measure the social benefit of an initiative against the financial investment required to deliver it. This happens to describe neatly the productivity task facing public service agencies: maximise your social benefit given a fixed investment of taxpayers' money.
In fact, governments have begun to apply these new approaches to traditional public sector tasks using social impact bonds. These instruments involve private and philanthropic investors investing their own capital to deliver public good outcomes. The first bond in the UK sought to reduce recidivism while three new bonds in NSW aim to assist vulnerable children, reduce the need for out-of-home care, and prevent adult reoffending.
These experimental new approaches are to be welcomed but they should not been seen as replacements for public sector capacity. If we continue to steadily shrink our public sector capacity, as has happened in NSW, QLD and VIC in recent years, we will lose thousands of years of accumulated know-how and experience. Should we decide that these cutbacks have been misguided, this know-how and experience will be impossible to replace.
Instead, governments should seek to experiment with these innovative new approaches within the public service. Per Capita proposes that governments trial social impact bonds with Treasury acting as the independent investor and groups of public sector staff acting as the agents of innovative change.
Under this model, public sector staff within a specific agency or departmental unit could opt-in to a 'public service venture'. This venture would agree to specific policy targets and a budget with Treasury and would be given a wide degree of operational autonomy to pursue these targets without intervention from above. Examples of such targets might include the on-time running of a district bus service, reducing obesity levels within a local population or speeding the recovery and return of injured workers to the workplace.
If the targets are met within the agreed timeframe, the venture's staff would receive incentive payments in excess of their usual salaries. The staff share in the risk of the venture too, so they face loss of income or other entitlements should the venture fail to achieve its targets.
This sharing of risk and reward is a central feature of the social impact bond model, and there is no reason why motivated public sector staff should not be able to access similar opportunities. If early trials of the public sector venture model are successful, it may evolve such that external investors are able to fund groups of staff that have earned greater autonomy through performance.
If, on the other hand, we fail to experiment with new approaches to unlock public sector innovation, we may one day find we have outsourced invaluable capability from the public sector that we can never replace.