Risks in social bonds

Insights Newsletter
5 June, 2015

Social bonds have caused quite a stir in the media this week. The model has attracted a deluge of negative feedback. Some of the criticism is warranted. Other concerns, however, can be effectively dealt with through careful contracting and design.
The government has announced the first social bonds pilot, which will focus on delivering employment services to people with mental health conditions.
Some have argued that social bonds are an internationally unproven model. It is true, social bonds are a new concept, not just in New Zealand, but around the world.
But it is a bit perplexing that social bonds have received such publicity when the Government is trying many different policy experiments in social development. Examples include Whanau Ora, Children’s Teams, Social Sector Trials, and Better Public Services targets. The social bonds model is just one of many.
Critics argue that the private sector would simply cherry-pick their clients who would most likely trigger a pay-out; that they would choose to take on the easiest cases (and reject the hardest cases) in order to make a profit.
Fewer people have pointed out the converse: that the government has an incentive to pass on their hardest cases to the social bonds programme, to avoid making a pay-out. Both are equally undesirable.
One method of dealing with this problem is to have willing participants enter a random lottery to be part of the social bonds programme. This would also mitigate the problem of self-selection bias in assessing whether the programme has worked.
Critics also highlight the risk of gaming: where people hit the target outcomes, but put their clients at risk. In other words, they meet the terms of the contract, but not its spirit. Again, careful contracting can target the intended outcome while monitoring for unintended consequences.
Both cherry-picking and gaming are concerns not just associated with social bonds, but most forms of performance-based contracting. Reassuringly, the government is already building expertise in this area.
The point of a pilot project, like the social bonds pilot, is to find out what works. As Eric Crampton mentioned above, this does not mean the government is experimenting on the mentally ill for profit.
As well as testing the model’s effectiveness, this pilot will offer rich lessons for future performance-based contracts, the collection of evidence-based outcomes, and alternative programmes to the status quo.

Stay in the loop: Subscribe to updates