Show simple item record

dc.contributor.advisorOwen, Dorian
dc.contributor.advisorHansen, Paul
dc.contributor.authorTrendle, Deborahen_NZ
dc.identifier.citationTrendle, D. (2004, December). Is an apple a day enough to keep the doctor away?: A time-series investigation into the determinants of selected diseases and aggregate mortality for New Zealand (Thesis, Master of Commerce). Retrieved from
dc.description.abstractIdentifying the determinants of health and ill-health in a nation has been the subject of an ongoing debate for several decades. Beginning with the work of Grossman (1972b), which was subsequently popularised by Wagstaff (1986), a substantial amount of economic research has been devoted to investigating the factors that enhance or diminish individual and population health. Previous literature indicates that numerous factors at both the micro and macro level can influence the degree of ill-health in a nation. They are generally referred to as social, economic, lifestyle, demographic, medical and government factors. Very little substantive research on this topic has been done in New Zealand to determine whether the results of overseas studies are applicable here. This study is an ecological investigation into the potential determinants of selected diseases in New Zealand, from 1950 to 2001. Using annual time series data, total mortality, and 13 different diseases are individually examined: Asthma, Tuberculosis, Malignant Neoplasm of the Colon, Breast and Lung, Diabetes Mellitus, Epilepsy, Hypertensive Diseases, Bronchitis Unqualified, Appendicitis, Acute Nephritis, Rheumatoid Arthritis and Osteoarthritis. The empirical analysis is conducted in two parts; initially models are developed which include current values of all explanatory variables and, secondly, variables appearing in the final models are lagged by ten years. However, as most of the data are nonstationary, the results of only those models that yielded evidence of cointegration were analysed. A generalto-specific modelling approach (using Hendry and Krolzig's (2001) PcGets algorithm) is applied to the general models, which include a broad range of potential health determinants. Error correction models are subsequently generated as a pragmatic robustness check for cointegration of the final model specifications. The final models for six diseases did not show evidence of cointegration and therefore, could not be investigated. The results suggest that the major determinants of ill-health and mortality in New Zealand are the legalisation of abortion (generally negative effects), average weekly wage rates (generally negative), female employment (generally positive), unemployment (negative), consumer price index (generally negative), marriage (generally negative), total and ex-nuptial births (generally positive), and health expenditure per capita (generally positive). Contrary to a majority of the theoretical and empirical studies, education, income inequality and various lifestyle variables appear to have a relatively insignificant effect across most diseases investigated.en_NZ
dc.subjectindividual and population healthen_NZ
dc.subjectdeterminants of ill-health and mortality in New Zealanden_NZ
dc.subjectgeneral-to-specific modellingen_NZ
dc.subject.lcshHC Economic History & Conditionsen_NZ
dc.subject.lcshR Medicine (General)en_NZ
dc.subject.lcshH Social Sciences (General)en_NZ
dc.titleIs an apple a day enough to keep the doctor away?: A time-series investigation into the determinants of selected diseases and aggregate mortality for New Zealanden_NZ
otago.schoolEconomicsen_NZ of Commerce of Otagoen_NZ Thesesen_NZ
otago.openaccessAbstract Only
 Find in your library

Files in this item


There are no files associated with this item.

This item is not available in full-text via OUR Archive.

If you would like to read this item, please apply for an inter-library loan from the University of Otago via your local library.

If you are the author of this item, please contact us if you wish to discuss making the full text publicly available.

This item appears in the following Collection(s)

Show simple item record