A joint Treasury, Ministry of Health release on the report, made public with redactions, said the main problem was a lack of commercial financial expertise within their pilot team. That led to a heavy focus on "process diligence" as opposed to achieving a successful outcome.
Other issues included a lack of visible senior sponsorship of the pilot within the agencies and the need to include investors earlier in the process and provide them with clearer information upfront on the basic commercial and financial parameters the Crown is willing to contract at.
At last month's Responsible Investment Association conference in Auckland, attendees speculated the government would need to pay at least a 12 per cent interest rate on the bonds to make it worthwhile for social investors. Social impact bonds have been in place overseas for some years and a report by think tank, The New Zealand Initiative, last year said they could markedly improve social outcomes and accountability in the social service sector.
Mules' report said the requirement to undertake a market-led process in New Zealand added time and complexity to the procurement process, along with a loss of Crown negotiating power, said the report. A lack of Crown "real-time decision-makers" at the negotiating table was also "sub-optimal".
Changes made to the bond pilot as a result of the review include clarifying roles and responsibilities between government agencies with the Ministry of Health remaining responsible for delivery and Treasury providing sufficient commercial advice, clearer delegation within Treasury, and changes to the membership and governance of the pilot team to include more senior "sponsors".
Mules' report recommended government should help provide proof of the concept by being willing to invest in impact bonds alongside other social investors and consideration should be given to government setting up a purpose-built social bonds investment vehicle.
The pilot phase of the programme will end when negotiations currently underway on the first two bonds close, regardless of whether they result in a contract, the joint release said.
The smaller mental health bond is relatively simple, "so the expectation is that contract could be reached", it said, while the larger second bond is at an earlier stage and requires more complex financial modelling. Outcomes of youth offending could be much later in life for recipients of the service - for example a 12-year-old receiving the service might then avoid a jail term in their 20s - which makes the financial modelling harder as the bond needs to pay out on demonstrated results that are more "proximate in time to the receipt of the service," it said.
Once the pilot has been concluded, the agencies will "formally capture lessons that can be applied to other forms of payments-for-results or outcomes-based contracting" and assess whether further work on other social bonds should continue. The release said it was too late given the "advanced stage" of negotiations for the first two bonds to include some of Mules' recommendations in the pilot and they would be best actioned in any future bond procurements.
