The Pitfalls of Monitoring Minority Labour Market Phenomena

Abstract

Labour markets are in a constant state of change, in both scale and composition. Policy analysts and researchers alike have an interest in tracking such changes over time. Statistics New Zealand is the major source of data on labour markets, and various statistical surveys and census keep us updated at varying frequencies, from quarterly to 5-yearly. For policy purposes. frequency and timeliness of monitoring data are important. The Household Labour Force Survey is therefore used as the statistical basis for monitoring various aspects of New Zealand labour markets, including trends in employment and unemployment levels. However, comparison of various statistical data sets reveals substantial differences in estimates as data are disaggregated, or as data refer to less universal (i.e. minority) labour market phenomena. Which data set is likely to be more accurate? Is reduced accuracy a casualty of survey sampling., weighting and estimation procedures?