11. 12. 13. 14. 15. 16. 17. 18. 19. 20. elevated in-work poverty rate relative to households of New Zealand European ethnicity (5.9 percent). Migrants: Households with at least one adult born in New Zealand face the average in-work poverty rate. There is substantial heterogeneity regarding in-work poverty prevalence across households with migrant adults, with those from North-East Asia experiencing one of the highest rates (+7.4 percentage points relative to sample average) and those from the United Kingdom experiencing a lower rate (-1.7 percentage points). Education: As expected, the in-work poverty rate is strongly associated with educational attainment of the household. The in-work poverty rate for households without any qualification is 10.7 percent – but it is only 4.9 percent for households with a bachelor’s degree as the highest educational level. Disability and health difficulties: Households with at least one disabled adult had a higher rate of in-work poverty (9.5 percent) compared to households without a disabled adult (6.6 percent). Health difficulties are also associated with higher rates of in-work poverty, particularly difficulties with learning (10.1 percent) and communicating/socialising (10.1 percent) compared to physical health difficulties (ranging from 7.9 to 8.5 percent). Household structure: The lowest in-work poverty rate is observed for households composed of a couple without children (4.8 percent), followed by a couple with child(ren) (6.3 percent) and single adults (6.4 percent). Households which are comprised of two or more families or single-parent households have elevated rates of experiencing in-work poverty (at 9.6 percent and 12.3 percent, respectively). Additional earner: Having a second worker in the household reduces the in-work poverty rate substantially. For example, for couples with children and only one adult working, the in-work poverty rate is 13.5 percent; this falls to 1.9 percent if there is more than one adult working. Occupation and industry: The in-work poverty rate is negatively associated with occupational hierarchy (an indication of a role’s level of educational requirements, social status, and income). The strength of this relationship is stronger if the occupation is associated with the main earner of the household. Furthermore, working in agriculture, forestry and fishing, and in the accommodation and food service industry is associated with elevated in-work poverty rates. Benefit receipt: Benefit receipt plays a greater role for single parents with children, for two or more family households and for one-person households relative to couples with or without child(ren). In-work poor households are more likely to be in benefit receipt (at least once in a year) relative to their in-work non-poor counterparts (28.7 versus 10.3 percent). Home ownership: The in-work poverty rate for homeowners is 5.2 percent, while this rises to 9.2 percent for renters. Moreover, where working poor households were renting, they spent (on average) every second dollar of their disposable income on rent, compared to every fourth dollar for in-work non-poor households. After housing costs: For renters, the in-work poverty rate is 12.8 percent after accounting for housing costs (based on self-reported rent costs in the 2013 Census). Poverty duration: Exploring within-year variation in poverty status, we find that as at March 2013, 20 percent experience at least one month of poverty in the twelve months preceding March 2013, and 1 percent experience poverty in all of the twelve months.