Unlocking the numerator-denominator bias. I: Adjustments ratios by ethnicity for 1991-94 mortality data. The New Zealand Census-Mortality Study

N Z Med J. 2002 Feb 8;115(1147):39-43.


Aim: To determine the extent of the under-reporting of Mäori and Pacific mortality among 0-74 year olds for the period 1991-94.

Methods: A subset (n=22,578) of highly probable linked 1991 census and 1991-94 mortality records were selected from the 31,635 census-mortality links in the New Zealand Census-Mortality Study. The numbers of decedents assigned as Mäori, Pacific, and non-Mäori non-Pacific were compared between mortality and census data.

Results: Compared to the death registration form, 29% more 0-74 year old decedents during 1991-94 had self-identified as sole-Mäori on the 1991 census (46% for prioritised-Mäori). This numerator-denominator bias was greater among the young and those living in central and southern New Zealand. Among 0-14, 15-24, 25-44, 45-64, and 65-74 year old decedents, respectively, 91%, 50%, 41%, 26% and 15% more decedents had self-identified as sole-Mäori on the 1991 census. For Northern, Midland, Central and Southern regional health authority areas, respectively, 14%, 17%, 81% and 102% more decedents had self-identified as sole-Mäori. Among Pacific decedents 68% more 0-74 year old decedents had self-identified as sole-Pacific on the 1991 census (78% for prioritised-Pacific group). This bias for Pacific decedents did not notably vary by age and region.

Conclusions: This study confirms substantial underestimation of Mäori and Pacific mortality rates for the period 1991-94, even using the recommended sole-ethnic group denominator. The results from this study should be used to adjust ethnic-specific mortality rates for the early 1990s. Population-based funding formulas that included region-specific Mäori mortality rates would have particularly disadvantaged central and southern regions.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Bias
  • Censuses*
  • Child
  • Child, Preschool
  • Ethnology / statistics & numerical data*
  • Female
  • Humans
  • Infant
  • Infant, Newborn
  • Male
  • Mathematical Computing*
  • Middle Aged
  • Mortality*
  • New Zealand / epidemiology
  • Pacific Islands / ethnology