Australian migration propensities by visa class: an analysis of linked administrative data

Abstract

Following a number of difficulties measuring net overseas migration (NOM), since 2006 the Australian Bureau of Statistics (ABS) has defined a usual resident of Australia as a person who spends at least 12 months of any given 16-month period in Australia. This is ascertained through matching of individual records for persons moving across Australia’s borders, including their Arrival and Departure Cards, their passport movements and their visa grants. The total number of persons counted into the population through arrival in a given quarter is termed NOM Arrivals. The total number of persons counted out of the population through departure in a given quarter is termed NOM Departures. The net of the two numbers is NOM or Net Overseas Migration for the quarter. In the late 2000s the data capturing NOM movements was held by the ABS and the data collected on visa grants was administered by the then Department of Immigration and Citizenship. In this paper, we examine the utility of these now linked administrative data sets in improving our knowledge about the timing and level of NOM propensities, or probabilities of ‘NOM tagging’ in the Australian migration system. Specifically, we present findings on the estimated event history of ‘NOMing in’ and ‘NOMing out’ of Australia using 22 aggregate visa types. These aggregate visa types are defined broadly containing categories such as business skills, employer sponsored, general skilled migration, working holiday makers, students and special humanitarian programme types. Implications of our findings with respect to improving Australian migration forecasting and further data linkages are discussed.