Thompson, V. Wesley2017-05-222017-05-222001http://hdl.handle.net/10179/11048This study attempts to fit the logit model to a random sample of data compiled by the New Zealand Employment Service on individuals who have completed unemployment spells, over the period 1988-1997. The objective is to estimate the probability that an individual job seeker, with a certain set of personal attributes, will become long-term unemployed. The regression results are consistent with a priori expectations. However, the predictive power of the model is low, lending support to conclusions from other empirical studies that have used other approaches to modelling long-term unemployment in New Zealand. That is, the current set of personal attributes on which data arc collected in New Zealand are inadequate for modelling long-term unemployment.enThe AuthorEconometric modelsUnemploymentNew ZealandHard-core unemployedThesisQ112857168https://www.wikidata.org/wiki/Q112857168