Veritas propter investigationem [Truth through research]
TitleDynamics of Income and Deprivation in New Zealand, 2002-2009 (2012)
AuthorsKristie Carter
 Fiona Imlach Gunasekara
InstitutionUniversity of Otago
TopicsLongitudinal survey
 Income mobility
 Statistical issues
CountryNew Zealand
Date Published2012
Date posted on PR02 Dec 2014
Carter, K Gunasekara, IF, (2012). Dynamics of Income and Deprivation in New Zealand, 2002-2009 (2012) University of Otago,

PensionReforms’ summary and comments

PensionReforms has previously reported on data obtained from New Zealand’s Survey of Family Income and Employment (SoFIE): see here, here, here, and here.


This 2012 report uses SoFIE to examine “the dynamics of low income and deprivation in New Zealand”.


SoFIE is a longitudinal study that started in 2002 and continued until the last tranche of data was gathered in 2010.  Because it is longitudinal, it can see “changes and trajectories” that the more common cross-sectional data miss.


For example, over a period of seven years many more people experienced low income than at one pointintime, where cross sectional low income (<60% of median household equivalised income) rates are around 24%.  However, the longitudinal estimate of low income over seven years was approximately double this (50%) – i.e. half of the sample experienced one or more years of low income.”


As expected, the proportions in this category were higher for the indigenous Maori group and those in sole-parent families.


Under the deprivation measures of poverty, (New Zealand Individual Deprivation Index score of three or more) the number in this category “…at any one time point was 67%, but the longitudinal estimate of deprivation over three time periods was almost twice this (12%).”


The longitudinal data can also show how much time people spent with low incomes or in deprivation.


The report also develops a measure of “chronic low income” (where smoothed average household income over the seven years fell below the average low income line over the seven years):

“Approximately two thirds of people who were in low income at any one point in time were chronically in low income, but this proportion was higher for Maori and children.”


This also meant that 30-40% of the population with those incomes did not experience it for the seven years “…meaning that their low income state was not persistent.”


A small group (5%) who were outside the low-income group at least once during the study period  “… were chronically in low income over the study period (and this was also higher for Maori and children), indicating that crosssectional measures of low income or poverty may underestimate the number of people in the population who are poor.”


These trends can be measured only through a longitudinal survey:

“We found high persistence of those in low income with about a quarter of respondents who were in low income at wave one being in low income for all seven waves.  There was also a lot of churn in entry and exit rates in and out of low income over the study period.  The twoyear entry rates into low income were around 7% and exit rates were 78%.


There is quite a lot of mobility in annual incomes with the most common movement being to an adjacent income quintile.


The report also found that the mean deprivation increased with the length of time that low incomes persisted.  But not all with low incomes reported deprivation and some outside the income measure did report deprivation.


“However, the results do not take into account changes in demographic events, such as forming partnerships, having children, marital dissolution, retirement or becoming employed, which have been shown to have an impact on income mobility and transitions in and out of low income over time.”


Future SoFIE modelling will take account of changes in family structure and employment over time.


PensionReforms supports the role of longitudinal surveys for work of this kind.  Cross-sectional surveys are too blunt an instrument and even with SoFIE the report concluded that more subtle measures of ‘low’ incomes and ‘deprivation’ seem needed.


Given that the objective of a state’s Tier 1 scheme should be to alleviate or even eliminate poverty in old age, PensionReforms supports this kind of research.  If nothing else, the state needs to know whether its policy on pensions is actually working.  (File size 610 KB; 48 pp) 707