Social investment is back – and so are the risks of using data to target disadvantage

With the recent establishment of a new Social Investment Agency – described as a “driving project” for the government by Finance Minister Nicola Willis – it seems New Zealand has come full circle on this approach to social welfare.

First championed by then finance minister Bill English in 2015, social investment was rebranded “social wellbeing” by Labour-led governments between 2017 and 2023. But Willis signalled before last year’s election that its time had come again.

In a speech in 2022, she argued taxpayer money wasn’t being spent responsibly by the Labour administration, and that a targeted social investment approach was needed. During the 2023 election campaign, the National Party promised social investment would return.

Essentially, the policy involves using data to calculate which groups of people cost the government the most over a lifetime. Interventions aimed at reducing that cost are then targeted at those people. The idea is that early investment saves later social costs.

Right now, however, we don’t know the finer details of how Willis intends to implement the policy. But we do know how it worked in the past – and what lessons might be drawn from its earlier, short-lived implementation.

An actuarial approach to welfare

In New Zealand, the idea of social investment can be traced back to the fifth National government which held office for three terms between 2008 and 2017.

In September 2015, English outlined his approach in a Treasury lecture, explaining how the government had commissioned Australian actuary firm Taylor Fry to calculate the lifetime welfare cost to the state of people on benefits.

Typically, actuaries use statistics to calculate risk for insurance companies, information that is then used to set premiums. English said the Taylor Fry calculations would identify which beneficiary “is going to cost us the most money”.

The answer was single parents receiving a benefit. Consequently, they were deemed most in need of direct government intervention, including giving an approved mentor control of their money.

According to English’s version of social investment, data enabled the government to calculate the “forward liability” of its citizens, and target interventions accordingly.

This is not the only way to define social investment, however, and other countries often adopt a more universal approach. For example, European models tend to focus on social equality and inclusivity rather than targeting specific groups.

English’s model focused on applying benefit sanctions and conditions. The aim was to “reduce the lifetime public cost of the welfare-recipient population, thereby offering fiscal returns-on-investment, absorbed into public coffers”.

A Social Investment Unit was created in 2016, followed by a Social Investment Agency in 2017. This was a standalone agency providing advice across government departments.

No accounting for structural disadvantage

Official thinking about social investment predates the establishment of the unit and agency. In 2015, the second of two reports produced by an expert panel review of the Child, Youth and Family agency (now Oranga Tamariki) recommended a new child-centred social investment agency be created.

The report’s analysis and advice focused on intervening early to reduce the risk of vulnerable children growing up to be beneficiaries, teen parents, substance users or prisoners (among other negative outcomes).

It was suggested these potential future behaviours almost always stemmed from the actions (or inactions) of parents. Māori were identified as being especially costly due to their over-representation in child protection statistics. They were described as a “forward liability associated with poor outcomes”.

The proposed response was early intervention and social investment. That would include the removal of very young children from whānau/families where they were perceived to be at high risk. The reasoning was that the predicted damage might then never eventuate, thereby saving taxpayer dollars.

As my doctoral research found, no consideration in the report was given to the effects of systemic conditions such as poverty and the legacies of colonisation.

Costs to the state

The social investment model, with its emphasis on financial liability to the state, became a major influence on Oranga Tamariki’s practice.

It led to an increase in the early removal of tamariki Māori, especially babies, from their birth families – as demonstrated in the 2019 Hawkes Bay “uplift” case, where social workers attempted to remove a Māori baby soon after birth.

In 2017, the new Labour government promised a review of the Social Investment Agency, renaming it the Social Wellbeing Agency in 2020. The social development minister at the time, Carmel Sepuloni, said the agency would have a more holistic approach. Data would be only one of a number of considerations when delivering social services.

But with the agency now reverting to its original name, the idea of using data to guide early intervention seems to be central again. It’s unclear, however, whether the actuarial approach of Bill English’s earlier model will return.

Nicola Willis does seem to be aware of the criticism of the English-era model’s apparent focus on fiscal risk and returns. She has stressed that measuring other outcomes is also important.

As yet, though, there is no indication the policy’s highly targeted approach to welfare will account for structural factors such as colonisation and poverty.

Given the government’s drive to remove any special policy considerations based on te Tiriti of Waitangi/Treaty of Waitangi, the risk remains that some Māori will again come to be viewed as a “cost” to the state.