Kieran Madden

By Kieran Madden - 21/08/2015

Kieran Madden

By Kieran Madden -

Like what you're reading?
Share it around.

Or just highlight the part you like...

Recommendations | Defining and measuring what it means to be poor in New Zealand


Defining Poverty: A definition of poverty cannot capture “everything in the world we don’t like,” as one roundtable attendee noted.[1] Instead, it must distinguish between those who are poor and those who are not.[2] Most agree that poverty involves a lack of resources that leads to exclusion from a minimum way of life.[3]

Recommendation 1: Poverty should be defined as a situation where: a person or family lacks the material resources to meet their minimal needs as recognised by most New Zealanders.

Expanding upon the key concepts in the definition:

  • Material Resources can be formal—that is, provided by the Government—or informal— provided by family, friends, neighbours, churches etc. There are two basic kinds:
    • Financial: Income, benefits, assets, material goods, charitable gifts etc.
    • In-kind: Health services, education, childcare from family, etc.
  • Minimal Needs are determined by what most New Zealanders consider necessary for a minimal acceptable standard of living to participate in society: a range of items or activities that no one should go without.[4] These needs may be social or material and go beyond what’s required for mere survival. The needs that are included are those that require material resources to fulfil (therefore the definition doesn’t include the full breadth of human needs like meaningful relationships, for example).[5] Needs change over time and differ depending on personal/family circumstances such as age, health, disability, geography, prices etc.[6]

Poverty is best understood as a dynamic relationship between these resources and needs. People use resources to meet their needs, and the scarring effects of deprivation (going without) and social exclusion (inability to participate in society) are likely to result when these needs are not met. MSD Researcher Bryan Perry’s stylised diagram above describes the relationship between resources, needs and other factors.

Some additional points of clarification on the recommended definition may also be helpful:

  • Lacking sufficient resources to meet minimal needs alone is what separates those who are poor from those who are not. This means that if people have sufficient resources to meet their needs but choose not to, they are not in poverty.
  • The definition does not (and does not need to) capture the various material and social causes of poverty: why people lack sufficient resources. It also doesn’t describe the consequences of going without: how their lives are blighted (deprivation or social exclusion).
  • While resources are primarily material, the causes and consequences of poverty can be social or spiritual: a lack of hope, ambition or motivation for example.
  • A person or family can be deprived or socially excluded without being poor.
  • Conversely, while prolonged periods of poverty are likely to lead to deprivation or social exclusion, this is not necessarily the case when people or families experience poverty for shorter periods.
  • A material resource-based definition like this does not necessarily lead to solely material resource- based solutions. For meaningful and sustained reductions in poverty, policy must tackle its causes and consequences alongside a focus on resource adequacy.

Measuring Poverty: Measures are signposts that point to poverty, necessary simplifications of a complex condition that are especially important for policy analysis. No one measure is sufficient as different measures tell different stories, serve different purposes and have different strengths and weaknesses. Some, like median income thresholds, provide an overall benchmark; others, like deprivation and outcome indicators, provide insights into the lived experience, causes and consequences, and which groups are most at risk of poverty. Both have distinct policy applications. Headline income measures can track broad progress of policy directions, while deprivation and outcome measures are much better suited to tailoring specific solutions based on clusters of needs in particular communities. Measures should be acceptable and understandable to the public, statistically defensible, comparable and consistent with the concepts used in the definition, and use data that are available or relatively easy to obtain.[7]

Recommendation 2: Regularly publish a poverty and deprivation dashboard including income measures, deprivation and outcome indicators.

Both income and deprivation measures should continue to be monitored and published regularly for households, including breakdowns by characteristics like family size, age, working status, housing type, ethnicity and so on where possible.[8] These are, at best, approximations of the numbers of those without sufficient material resources to meet their minimal needs. Ministry of Social Development’s (MSD) annual Household Incomes in New Zealand report is excellent in this regard and should continue to be supported. Poverty-related outcome indicators should also be regularly monitored and published.[9] We broadly commend the Children’s Commissioner’s Expert Advisory Group’s (EAG) recommendations on poverty measurement.[10]

The dashboard below features five income and deprivation measures/indicators, and around ten poverty-related outcome indicators:[11]

Recommendation 3: Use consensual budget standards to identify what most New Zealanders think is a minimal acceptable standard of living and potentially derive an income threshold from this process.

MSD’s submission highlighted the need to “identify what ‘ordinary New Zealanders’ see as an unacceptably low level of income or material wellbeing.” We agree. Rather than relying on a proportion of people falling below median income thresholds as a strong indicator of insufficient resources, the consensual approach identifies what the minimal acceptable standard looks like by public consensus through surveys and focus groups mediated by experts. A threshold derived from this would be more democratic, methodologically consistent with our definition, and potentially more acceptable and meaningful to the public. Done well, this should be added to the dashboard as a complement or even as a replacement for the income thresholds.

Robust work from the UK could be replicated here. The Joseph Rowntree Foundation’s Minimum Income Standard uses focus group discussions to specify a basket of goods and services necessary to arrive at “the income that people need in order to reach a minimum socially acceptable standard of living in the UK today, based on what members of the public think.”[14] Poverty and Social Exclusion UK uses a survey to understand what the population deems are necessities, and “identifies people falling below what the public agrees is a minimum standard of living” using deprivation measures.[15] While not consensual, Canada’s Market Based Measure uses a basket of goods approach to complement income measures.[16] Pioneering focus group work was done in New Zealand in this area, however, results need to be updated to reflect current conditions.[17] This need not be implemented entirely by government but should be commissioned and supported by it; a university department could develop it for example.

Recommendation 4: Use clustering statistical techniques to target, tailor and evaluate policy by identifying people and families with different combinations of risk factors.

As both Treasury and MSD have noted in recent reports, “cumulative impact of multiple disadvantage across different domains” is more likely to lead to poor outcomes and opportunities than low income alone.[18] Those in poverty are not one homogenous group. Clustering techniques like latent class analysis or factor analysis provide a richer picture by grouping families or households into clusters that share similar outcomes. Groups identified this way go beyond family size, age, working status, housing type, ethnicity etc. UK think tank Demos’ work Poverty in Perspective is exemplary, and aims to “prompt more holistic and multi-agency solutions (based on an understanding of multiple factors) regarding how each group might be helped out of the distinct type of poverty they face.”[19] Treasury has investigated use of these techniques in a working paper,[20] as has the Growing Up In New Zealand study in their Vulnerability report.[21] Understanding these groupings is key to targeting, tailoring and evaluating policy responses more effectively. This need not be implemented by government but should be commissioned and supported by it.


Recommendation 5: There should be some legislative requirement that the measures and indicators above are regularly published.

There should be some requirement that these measures are made available to the public at regular intervals. Legislation like that in the Families Commission Amendment Act 2014 that requires the Commission to develop and publish “an annual Families Status Report that measures and monitors the wellbeing of New Zealand families” should be considered.[22] Ideally, Statistics New Zealand would take primary ownership of this requirement as an independent agency with relevant expertise.

Recommendation 6: A poverty-specific legislative framework should not be implemented.

While many, including the EAG, have advocated for a Child Poverty Act to incentivise and coordinate policy action, we believe that the democratic process as it stands provides sufficient means for citizens to signal their preferences and policy priorities to government. Not only would enshrining a legislative framework be extremely costly to implement, similar legislation in the UK has proven to be ineffective at reducing levels of poverty.[23] The intended benefits could instead be realised by harnessing and expanding existing, less rigid mechanisms such as the Results for New Zealanders framework.

Recommendation 7: Extend the Better Public Service targets / Results for New Zealanders framework to include reasonable, time- specific targets aimed at reducing poverty.

Thoughtfully-set targets would signal governmental priorities and aspirations, galvanise action across agencies, and provide an additional accountability mechanism. Reasonable targets that align with the above outcome indicators should also be considered, alongside pre-existing targets such as long-term benefit dependency and rheumatic fever. It should be made clear that the Government doesn’t have complete control over movement towards targets (as with any complex social problem), but it should be required to outline policy strategies towards reaching the set targets to explain performance based on evaluations of the effectiveness of these policies. The current medium-term five year horizon is appropriate.


Recommendation 8: Further investment is required in better data sources in New Zealand, particularly longitudinal studies like SOFiE, to understand the causes, consequences and dynamics of poverty.

Rigorous studies of the relationship between income poverty and material deprivation measures over time show that while even short periods of poverty can scar (house sale, debt, family breakdown) it is “people’s underlying economic position that matters, not their short-term fluctuations from year to year.”[24] Just as brief periods of poverty are unlikely to have long-lasting effects, long-term chronic poverty results in severe deprivation, so seeing permanent economic changes for families is important.

Recommendation 9: Official datasets should be more easily accessible to researchers.

From submissions received and discussions with experts and academics, the message was clear that socio- economic data in New Zealand is very difficult to access, prohibitively expensive or both. While data security should remain paramount, an investigation into improved data accessibility processes based on international best practice should be undertaken.


This set of recommendations on defining and measuring poverty is the first step in our journey towards developing and advocating for policies that give struggling New Zealanders the help they need and deserve. Informed through our research and consultation process, we now have a more robust understanding of what poverty is. This is a good start but it doesn’t change lives. For truly effective policies, we need a deeper understanding of the underlying causes and the damaging consequences of poverty. Our next piece of work will investigate these, with our findings to be released later in the year.

This is an extract from Kieran’s research series “The Heart of Poverty | Defining and Measuring What it Means to be Poor in New Zealand” Policy Paper. (Released 2015) 




[1]While a broader view on improving well-being is a legitimate policy goal (like that of Treasury’s Living Standards Framework ( abouttreasury/higherlivingstandards)), focusing on poverty requires a tighter definition.
[2] In real life, however, poverty exists as more of a continuum. Those just above a poverty line are not in a significantly different situation to those just below, for example.
[3] R. Berthoud, M. Bryan & E. Bardasi, “The dynamics of deprivation: the relationship between income and material deprivation over time,” (Department of Work and Pensions, 2004).
[4] See B. Perry, “Household incomes in New Zealand: Trends in indicators of inequality and hardship 1982 to 2013,” (Ministry of Social Development, 2014), 115 for a list of basic material needs. See also Recommendation 3 for more on determining basic needs. Ideally, these would be determined using a consensual method.
[5] For a list of human needs / dimensions of well-being, see Appendix 1 in K. Madden, “The Heart of Poverty: Matching passion with precision”, 30.
[6] Research from the UK suggests that while public perceptions of necessities usually rise, they can fall too. For example, “in all previous surveys over the past thirty years, being able to afford to give presents to family and friends once a year (such as on birthdays or at Christmas) was considered to be a necessity by the majority of people. In 2012, the majority of people no longer believe this is a necessity. The minimum expectations of the population have fallen.” See D. Gordon et. al, “The Impoverishment of the UK – PSE UK first results: Living Standards” (2013), 6.
[7] R. Michael & C.F. Citro, eds., “Measuring Poverty: A New Approach, report for the Joint Economic Committee of Congress” (National Academies Press, 1995).
[8] Different equivalence scales—adjustments to incomes take into account economies of scale within a household—can result in significantly different levels of poverty or hardship at a point in time. Over time, however, the trends are similar. The most appropriate method for New Zealand would need to be considered for the measurements recommended in this paper. See B. Perry, “Household incomes in New Zealand: Trends in indicators of inequality and hardship 1982 to 2013,” 236-240.
[9] B. Perry, “Household incomes in New Zealand: Trends in indicators of inequality and hardship 1982 to 2013.”
[10] Children’s Commissioner’s Expert Advisory Group, “Solutions to Child Poverty in New Zealand evidence for action” (Office of the Children’s Commissioner, 2012). For benefits and drawbacks of different measures, see K. Madden, “The Heart of Poverty: Matching Passion with Precision,” 16-23.
[11] Table adapted from Expert Advisory Group’s recommended suite of child poverty measures, Expert Advisory Group, “Working Paper 1: Defining and Measuring Child Poverty” (2012), Table 1, 7.
[12] Otago University’s NZDep indicators work is an effective complement to the MWI that shows geographical characteristics at the mesh block level. Geographic data helps target policy and should continue to be supported and expanded where possible. University of Otago Department of Public Health, “Socioeconomic Deprivation Indexes: NZDep and NZiDep,” (accessed 9 December 2014).
[13] See Recommendation 8 on longitudinal data sources.
[14] A. Davis, D. Hirsch & M. Padley, “A minimum income standard for the UK in 2014” (Joseph Rowntree Foundation, 2014), 8.
[15] D. Gordon et. al, “The Impoverishment of the UK – PSE UK first results: Living Standards,” (2013).
[16] Statistics New Zealand, “Measuring child poverty in New Zealand: Issues and practicalities, Paper prepared for the Expert Advisory Group on Solutions to Child Poverty,” (2012), 9.
[17] R. Stephens, C. Waldegrave & S. Frater, “Measuring Poverty in New Zealand,” Social Policy Journal of New Zealand, No.5 (1995).
[18] New Zealand Treasury, “Improving outcomes for children –Initial Views on Medium-term Policy Directions, Report to the Ministerial Committee on Poverty,” (2013),
B. Perry,” Household incomes in New Zealand: Trends in indicators of inequality and hardship 1982 to 2013,” 10.
[19] C. Wood et. al, “Poverty in Perspective,”(Demos, 2012), 14.
[20] See C. Ball & M. Ryan, “New Zealand Households and the 2008/09 Recession, New Zealand Treasury Working Paper 13/5,” (New Zealand Treasury, 2013).
[21] S. Morton et. al, “Growing Up in New Zealand Vulnerability Report 1: Exploring the Definition of Vulnerability for Children in their First 1000 Days,” (Growing Up in New Zealand, 2014).
[22] Families Commission Amendment Act, New Zealand Statutes, (2014).
[23] J. Hancock, “Legislating to Reduce Child Poverty” (Office of the Children’s Commissioner, 2014), 13-15.
[24] R. Berthoud and M. Bryan “Income, deprivation and poverty: a longitudinal analysis.” Journal of Social Policy 40, no.1 (2011): 135-156.

Post Tags:
Kieran Madden

By Kieran Madden -

Like what you're reading?
Share it around.

Or just highlight the part you like...

Want to know more about Maxim Institute and what we do?

Find out more


FORUM (monthly eNews)Event Invitations