The decile debate continues

The NZ Herald today broke a story that the Ministry is still intending to ‘abolish’ school deciles. Once unpicked, the story is slightly different. Essentially the proposal (whose source is not clear) is to shift from the current 5 indicators (based on the census) to 4 new indicators (measured through data-sharing between government agencies) for the purposes of determining additional school funding to overcome known disadvantage.

Leaving aside for a moment the abolition of the decile system (which of course is only a ranking of schools according to the current measurement data), the proposal is therefore: “We are changing the indicators by which school disadvantage is measured”.

The current census measures used are: low income household, no parental qualifications, low skill occupation, benefit-led household and household crowding (defined as more than two people per bedroom).

There is an emphasis in the current model on income and skill factors. I have said elsewhere that the decile system mirrors child poverty, in that most of the bottom forty percent of schools are full of children living in households below the poverty line. In a sense, the current model is a proxy for child poverty factors.

As around 85% of school achievement can be predicted by such factors, it is a reasonably good basis for additional school funding. Most countries use similar indicators to determine extra funding.

In a 2014 study by the University of Otago, it was suggested that there should be even “greater weight on low household income”, as the best predictor of future qualification change (i.e. the more funding based on low household income, the better the improvement).

The proposed new factors suggested in the information provided are parental imprisonment, child abuse, benefit and mother’s qualifications.

The latter two factors are similar to two factors in the current model. But any indicators of low working income, occupational skill or household crowding are to be replaced, the proposal reads, by two other indicators: imprisonment rates and child abuse. This goes directly against the Otago findings.

As the author of the only research ever to be completed in New Zealand on the relationship between parental imprisonment and child educational outcomes, I am pleased that my finding of poor education outcomes for these children has been taken seriously. But, while the effect is clear, I would not have thought that this is an indicator that is well-enough developed to fund schools. I can see at least three problems:

1. The average prison sentence is very short. Some people go in and out of prison on many occasions, others only have one sentence. Will any term of imprisonment do? What is the difference in effect between the father and the mother going to prison?
2. Many children of prisoners are invisible because their imprisoned parent does not appear on the birth certificate. Some have multiple ‘fathers’ and more than one ‘parent’ in prison. This complicates things.
3. I am a expert in this literature and I am yet to see a clear enough definition of the harm caused by parental imprisonment on a child’s education. Knowing that harm is caused is not the same as basing funding on it.

In short, while my research demands additional support for the children of prisoners, I am yet to be convinced that it is a robust enough indicator for school funding purposes.

I am not an expert in child abuse, but it seems reasonably clear that many of the same problems exist if we are to turn this into an indicator for funding purposes. What kinds of abuse will count? What effects will frequency and severity have?

The other issue that needs to be canvassed is questions around privacy and confidentiality. Many parents turn to their schools for support when terrible things like imprisonment and child abuse happen. Many others do all in their power to keep such information from schools, as it can be used to stigmatise families and discriminate against children. It is implied that the schools will know, at least, how many children fall into these categories, if funding is to be based on them.

The data-sharing that allows Ministries to know that certain factors affect learning occurs through Statistics NZ using anonymous identifiers (everyone has a unique number, but absolutely confidential). There will be nothing anonymous about the model if it is used to fund schools. At the least, the school will know it has 46 children of prisoners or 33 victims of child abuse enrolled. This needs MUCH more work. But, actually, I think the problems I have outlined are basically insuperable, and that, whatever funding model is developed, the indicators of parental imprisonment and child abuse will not eventually be incorporated.

Finally, any set of indicators that are developed can be groups into decile or quintiles by ranking them from most support needed to least. This is not the abolition of the decile funding system, but new names for old problems.

A footnote: The media should bear in mind that, according to the OECD PISA in 2009, the clearest indicator of learning outcomes in NZ is the number of books in the home, with those with fewer than 10 books are some 3.5 years behind in their learning at age 15 compared to those with 300 plus books. Why not use that indicator?

ENDS

© Scoop Media