Saturday, May 15, 2021

Defining ethnicity for the purposes of measuring child poverty

The Child Poverty Related Indicators report on housing, food security, school attendance and potentionally avoidable hospital admissions. The problem - or absurdity - is that each indicator uses a different method of defining child ethnicity.

• Housing affordability and housing quality indicators: ethnic groups are reported using the total response method. People were able to identify with more than one ethnic group; therefore, figures will not sum to the total population. People who responded ‘New Zealander’ were classified as ‘other ethnic group’.

• Food security: respondents can identify with multiple ethnicities, including specifying an ethnic group not listed.

• School attendance: students were able to identify with up to three ethnicities. Students are counted under each ethnic group they identify with and once in "Total".

• Potentially avoidable admissions: The ethnicity reported is based on prioritised ethnicity for a patient within the hospital inpatient system. This prioritises people to Māori, then Pacific, and then Other ethnicities. A person identified as having more than one ethnicity will be prioritised accordingly and will only be counted once under each ethnicity category for each PAH event.

Take the first, housing. If a child falls in the housing hardship group, and has two or more ethnicities, he is counted twice or more. So there is a bias towards overstating a housing hardship problem for children of mixed ethnicity. 

Conversely, with hospitalisations, even though each child is counted just once, because Maori ethnicity is prioritised, there is a bias towards overstating the Maori aspect of hospitalisation. 

Just two more reasons why the Child Poverty Related Indicators need to be taken with a grain of salt.

No comments: