Australian policymakers know well the cost of dealing with social crises is far greater than prevention, but structural factors make it difficult to intervene effectively. Peter Willis and Matt Tyler on bringing governments and communities back on the same page.
Crisis. Four Corners expose. Media flurry. Public shocked. Cue Royal Commission, Inquiry, or Review. Pressure builds and then subsides. Another crisis. Repeat.
This is too often the story of Australian social policy. Hard working governments and service providers responding to crises with few structures that direct focus towards effective prevention. We’ve seen “shocking and systemic failures” at the Don Dale Youth Detention Centre, ongoing tragedies in aged care facilities, and growing concerns as emergency rooms fail to deal with acute mental health episodes.
In July, we wrote an article in the Sydney Morning Herald proposing that the Australian government create a Social Outcomes Fund, a world first funding mechanism to increase the focus on preventative social programs.
Under a Social Outcomes Fund, the federal government would set aside funding to invest in preventative social programs that seek to avert social crises (e.g. homelessness or prison). Program effectiveness would be evaluated independently and, subject to meeting targets, the parts of government that benefit would replenish the Fund so that it can make more investments in the future.
A number of policymakers responded to our article, raising questions and asking for additional detail. We have written this article to address these questions and further the debate.
What’s the problem that the Fund is trying to solve?
Government spending in social services is skewed towards dealing with the consequences of crises, rather than preventing them from happening in the first place.
Consider, for example, that over the last five years, spending on prisons (crisis) has grown faster than spending on schools (prevention). Across Australia, less than 10% of child protection budgets go towards intensive family services – programs designed to prevent children from being separated from their primary caregivers. Departments responsible for child protection are often siloed from the departments with control over the underlying issues that drive child safety concerns – substance use, mental illness, and domestic violence.
Limited assessment of program effectiveness goes hand-in-hand with the lack of focus on prevention. For instance, the recently released Independent Review of Out of Home Care in New South Wales (the “Tune Review”) found 67% of programs targeting vulnerable children and families – worth $302 million – have not been evaluated. Where evaluations do exist, the Tune Review found “lack of outcomes information can in some cases compromise the evaluation”.
We believe that the lack of focus on effective prevention in social services is caused by four structural factors that are addressed by the Social Outcomes Fund.
First, many preventative programs take 10-15 years to show results. This is well beyond the timeframe of a normal electoral cycle. The Fund would have an explicit mandate to invest over a 10-15 year timeframe.
Second, benefits from preventative programs are often diffused across government departments. An effective early childhood education program will both benefit participants and create savings for multiple state (corrections, schools) and Commonwealth (health, welfare) government departments. While these benefits are spread across budgets and layers of government, program expenditure is often entirely borne by a single department. For this department, expenditure may exceed benefits. The Fund breaks down these silos and allows the costs and benefits of a program to be assessed on their merits – not based upon where they accrue.
Third, it is politically difficult to move money away from spending that is responding to acute crises (e.g. children in out-of-home care). Initially the Fund would be financed using new money, meaning that money would not need to be diverted away from crisis response. Subsequently, the Fund would be replenished using cost savings subject to achieving agreed targets (see “Who would pay for it?” below for detail).
Fourth, under traditional grant funding, it’s possible for governments to spend money without a mechanism to understand effectiveness. Even when evidence is available, often it does not guide funding allocation. For example, JPAL – a research institution in the US focussed on policy evaluation – has supported 626 evaluations since its inception but there has been only 15 instances where government chose to scale up a program that it found to be effective. This finding highlights the importance of linking Fund replenishment to outcomes. Doing so means there are increased incentives to generate rigorous evidence and allocate funding based on the evidence.
What kind of programs would it fund?
The Fund would support preventative programs that seek to improve a small number of measurable social outcomes. These programs would be additional to existing services and would be backed by evidence.
Example target outcomes could include decreasing the number of children requiring out-of-home care, decreasing crime levels, decreasing obesity and/or diabetes. In each case, achieving the outcome would result in cost savings which would be used to replenish the Fund. The target outcomes would be publicly communicated and tracked over time.
Who would pay for it?
The money for the Fund would come from the federal government. Based on similar initiatives elsewhere and with the view to expanding over time, we suggest approximately $100 million to start with.
If the evaluation of a program supported by the Fund found that it had created savings for other government departments, those departments would “replenish” the Fund. For example, if a program supported by the Fund decreased re-offending, costs savings would be obtained across the prison, policing, court, and welfare budgets. A portion of the savings across these budgets would be used to replenish the Fund.
Many social services are delivered by state governments who would participate on an “opt-in” basis. States who “opt-in” would receive upfront funding from the Fund and negotiate replenishment terms with the federal government based on the relationship between target outcomes and estimated costs savings (e.g. how a decrease in re-offending impacts the number of nights prisoners spend in jail).
To gain early momentum and engender focus, performance could be monitored on a annual basis with interim replenishment payments (prior to the 10-15 year mark) based on leading indicators. For instance, in the case of homelessness, a small interim payment could be made if someone enters permanent housing with longer term payments based on long-term stability and well-being. Funding could even be pulled early if, for example, program enrolments are too low. Interim measures could be used to adjust the program and associated systems (e.g. service referral pathways) to pursue improved performance.
To guard against the risk that state governments don’t make replenishment payments, there could also be the implied threat of reducing (related or unrelated) future Federal Government funding allocations to state governments.
Who would determine what the Fund invests in?
We envisage the Social Outcomes Fund being embedded in the federal government, allowing for national impact.
The target outcomes of the Fund and reporting requirements would be controlled by the federal parliament and would be legislated by Members of Parliament.
Selection of specific projects would be done by an independent board appointed by the Finance Minister or Treasurer. This would be similar to the way that the Treasurer appoints the Reserve Bank’s board. The independence of the board is important to avoid political influence and “pork barrelling”.
The board would be supported by a secretariat within the Department of Finance or Treasury who would approach preventative investments like an impact investor, using available evidence to model likely outcome improvements and cost savings scenarios. The secretariat would provide this analysis to the board to assist them when making their investment decisions.
How does it relate to Social Impact Bonds and other innovative government financing ideas, here and overseas?
The Social Outcomes Fund draws on the lessons learned from a different way of financing preventative programs: Social Impact Bonds. SIBs are a way to tie payment for services to measured delivery of outcomes. SIBs have been adopted especially in the US, UK and Australia and have been used in a variety of policy areas, including prisoner re-offending, homelessness, maternal health, preventing entry into out-of-home-care, and improving employment outcomes. Specific funds have been set up overseas (e.g. Life Chances Fund in the UK) to promote SIBs.
Although they have played an important role in promoting government effectiveness, there are a number of challenges with SIBs which the Social Outcomes Fund is designed to address. To date, SIBs have been relatively small; the largest project in Australia is $50 million over 7 years, compared to annual social sector expenditure of approximately $400 billion, excluding income support. Part of the reason for the lack of scale is the need to raise private capital for a new and unusual financial product. Private investors also add significant complexity and cost due to diverse motivations as well as commercial pressures to obtain a decent return. As a result, SIBs have been slow to put together, often taking over a year to negotiate.
The Social Outcomes Fund aims to take the best part of the SIBs model – the concept of government paying for successful preventative outcomes – and scale it up. In short, the Fund does not need private investors so can get to scale quicker.
What are the risks and how do we mitigate?
There are several risks associated with establishing a Social Outcomes Fund. We believe they can be mitigated with careful design.
First, there is the challenge of goal dilution – without a clear focus, the Fund could support a lot of small programs that would not actually improve outcomes. Having a small number of high impact and measurable public targets for the Fund would help mitigate against this happening.
Second, the time horizon for success (10-15 years) may make the Fund less politically palatable. Some politicians may consider achievement of desired outcomes as too distant. To mitigate this, the Fund could release real-time data (e.g. total number of acute mental health hospital admissions) related to the programs it is funding so that policymakers and politicians could track program progress. Future benefits and liabilities would be clear. Interim payments could also be made based upon annual leading indicators (see “Who would pay for it?” above)
Third, there is a risk that programs fail to prevent negative social outcomes. The reality is, when rigorously evaluated, many social programs have no or limited impacts. Expectations of both the public and associated stakeholders should be managed so that results are understood in context and that early failures do not place the Fund at risk. Relatedly, messaging should avoid promising that even successful programs will “pay for themselves”. Instead emphasis should be placed on combining fiscal responsibility with an emphasis on improved lives.
How would we know if the Fund is working?
Progress towards the Fund’s target outcomes would be tracked and reported over time. Cost savings associated with achieving targets would also be monitored. In New Zealand, for example, both sides of politics have agreed to report on future liabilities associated with any kind of accident (through the Accident Compensation Corporation) and the long-term forward costs of the welfare system (through the Ministry for Social Development)
Beyond examining outcomes, the rigour of the Fund’s evaluations and operations could be reviewed regularly by the Australian National Audit Office. System re-engineering among governments and service providers could be carefully monitored. Initiatives similar to the Fund overseas, such as the UK’s Troubled Families Programme, have created a variety of system changes including improved matching of people to services and the establishment of mechanisms to monitor data in real-time.
The impact of the Fund on other government expenditure should also be monitored. It’s unlikely that the Fund will be able to directly transform social services. Instead the Fund should seek to influence funding allocation elsewhere. Successful programs supported by the Fund should be scaled-up. Conversely, if programs supported by the Fund are not effective, spending on similar programs should be carefully examined and potentially reduced (noting that context, as well as implementation, impacts effectiveness).
Who gets the credit if it works?
The Fund would involve a wide range of ministers in its operations, ensuring that the credit for success would be shared across the federal government.
If the Fund was successful, the Finance Minister or Treasurer as head of the Fund would get a significant amount of credit.
Federal ministers who have responsibility for the policy areas where the Fund is targeting improvement (e.g. Health Minister if the Fund is targeting reducing obesity) would also share the success.
At the state level, relevant ministers and premiers could claim credit for opting-into the Fund. Cost-benefit ratios for each state could be calculated and communicated regularly.
Ultimately, those involved in the Fund would be establishing a social infrastructure legacy that could transform the way governments in Australia fund services. We must move beyond merely reacting to crises and developing social policy by Royal Commission. The Social Outcomes Fund is an important step in making the transition towards effective prevention of many of the challenges facing our society.
Peter Willis and Matt Tyler are graduates of Harvard’s Kennedy School of Government. They have worked for governments in the United States and Australia to improve social services including across child welfare, homelessness, criminal justice, education, and health.