This paper examines whether there would be a difference in the nationally reported proportions of each ethnicity within the New Zealand (domestic) school population if all ethnicities collected at the school level for each student were considered. At present the sizes of ethnic groups are reported by the Ministry of Education using a system that Statistics New Zealand used to use but no longer recommends where individual responses are prioritised so that only one ethnicity per person is counted.

National ethnic data for 10 to 14 year-olds from the schools’ 2006 Roll Return are compared firstly with total response data from the 2006 New Zealand census, and then with data from that census that have been prioritised to conform with the Roll Return system.

These data comparisons show that prioritising data does not give a full representation of the ethnic population of our schools. The ethnic make-up of our school student roll, and the ethnic representation given in education indicators, will only be fairly represented when total counts of all self-identified ethnicities are reported.