|dc.description.abstract||This study examined the new company failure rates of New Zealand's three main urban centres (Auckland, Wellington and Christchurch). The populations of newly registered companies within the three main urban areas in 1987 were used and consisted of 5413 companies in Auckland, 2219 companies in Wellington and 1235 companies in Christchurch. To allow for greater validity two measures of failure were used. These measures differ in their concentration and were discontinuance for any reason and liquidation.
The Auckland, Wellington and Christchurch company populations were treated as a representative sample of New Zealand companies, and a national new company failure rate was calculated. Along with the national company failure rate, regional company failure rates were calculated and compared using survival analysis techniques. The log-rank, Breslow and Tarone-Ware tests for equality were computed providing evidence of small levels of variation in new company failure rates between the three urban areas. Based on the measure discontinuance for any reason, Auckland registered companies had the best survival chances, Wellington registered companies had the second best survival chances and Christchurch registered companies had the worst survival chances. Based on the measure liquidation, however, Christchurch registered companies had the best survival chances, Wellington registered companies had the second best survival chances and
Auckland registered companies had the worst survival chances.
The Cox regression model with time-varying covariates was used to measure the explanatory power of the five independent variables (Consumer Price Index, unemployment, average weekly wage, average house price and population). From this a final model was constructed consisting of three variables that effect regional variation in new company failure rates (unemployment, Consumer Price Index and average house price).
The findings of this study confirm previous concerns relating to the deficiency of database management by the Ministry of Commerce Companies Office. It was recommended that the procedure used in this study be replicated periodically and the results presented in the form of a regional risk index. Other limitations of this study are discussed along with the identification of directions for further research.||en