Skip to main content
PMC home page
SSM - Population Health logoLink to SSM - Population Health
. 2021 Jun 8;15:100842. doi: 10.1016/j.ssmph.2021.100842

Service usage of a cohort of formerly homeless women in Aotearoa New Zealand

Brodie Fraser a,∗, Maddie White b, Hera Cook a, Elinor Chisholm a, Jenny Ombler a, Saera Chun a, Hiria Tareha c, Nevil Pierse a
PMCID: PMC8209275  PMID: 34169140

Abstract

Purpose

The aim of this paper is to explore government service usage across the domains of health, justice, and social development and tax for a cohort of formerly homeless people in Aotearoa New Zealand, focusing specifically on the experiences of women. The Integrated Data Infrastructure is used, which links our de-identified cohort data with administrative data from various Aotearoa New Zealand Government departments.

Results

Of the cohort of 390, the majority (53.8%) were women. These women were more likely to be younger (57.1% were aged 25–44), indigenous Māori (78.6%), and have children (81.4%). These women had lower incomes, and higher rates of welfare benefit receipt, when compared to men in the cohort and a control group of women from the wider population.

Conclusions

The cohort were primarily female, younger, Māori, and parents. They earned much less than their non-homeless counterparts, and relied heavily on government support. The neoliberalisation of the welfare state, high rates of women's poverty, and the gendered nature of parenthood means that women's homelessness is distinct from men's homelessness.

Keywords: Homelessness, Housing first, New Zealand, Women, Linked data

Highlights

  • •

    The cohort had distinct government service usage when compared to a cohort of homeless men, and non-homeless women.

  • •

    The cohort were more likely to be Māori, younger, and parents.

  • •

    The cohort required greater levels of income support than homeless men and non-homeless women.

1. Introduction

This paper provides a quantitative exploration of the service usage of a cohort of homeless women1 in a small city in Aotearoa New Zealand. It shows that homeless women are more likely to be younger, Māori, and parents. It builds on prior work conducted by He Kāinga Oranga (Pierse et al., 2019).. The experiences and needs of homeless women remain under-researched, despite a growing effort to address this disparity of information (Bretherton, 2017; May, Cloke, & Johnsen, 2007; North & Smith, 1993; Phipps, Dalton, Maxwell, & Cleary, 2018; Pleace, 2016; Reeve, 2018). The focus is usually on single men who are rough sleeping and are often dealing with addiction or poor mental health (Hagen & Ivanoff, 1988; Phipps et al., 2018). Definitions of homelessness derived from male experience has excluded the types of homelessness that many women experience (such as staying with friends and family), as well as the impact of domestic violence, motherhood, and responsibility for children (Bretherton, 2017; Pleace, 2016). Women's experiences of homelessness are different, and must be treated as such; comprehensive research is needed to explore this diversity.

The aim of this paper is to investigate gender differences in a subsection of the Aotearoa New Zealand homeless population. This is done by analysing service usage prior to being housed by a cohort of formerly homeless people who have been re-housed by Housing First (HF) services. De-identified and integrated administrative datasets have been used in this analysis. Housing First is an holistic approach to addressing homelessness, which is premised on the idea that complex issues are best addressed from the starting point of permanent housing (Tsemberis, 2010). This contrasts to more traditional models of addressing homelessness in which sobriety or other requirements must be met in order for clients to obtain and maintain housing (Pierse et al., 2019). This paper builds on previous research by He Kāinga Oranga, which showed that previously homeless people who were housed by The People's Project (TPP), a HF provider in Aotearoa New Zealand, had consistently higher rates of government service usage over a long time period prior to being housed (Pierse et al., 2019). 53.8% of the cohort of 390 discussed in this earlier research—and further investigated here—were female (n = 210), compared to 46.2% who were male (n = 180) (Pierse et al., 2019). This paper expands on this original analysis to explore the gendered experiences of homelessness amongst our HF cohort.

There is limited published research that focuses specifically on homeless women's lives in Aotearoa New Zealand and that which does exist is qualitative (Bukowski & Buetow, 2011; Groot, Hodgetts, Waimarea Nikora, & Leggat-Cook, 2011). Both quantitative and qualitative research is needed to fully understand women's homelessness. Rates of homelessness in Aotearoa New Zealand have been steadily growing over the past two decades, with these rates highlighting disproportionate experiences of homelessness by Māori, Aotearoa New Zealand's indigenous people (Amore, Viggers, & Howden-Chapman, 2020). Aotearoa New Zealand is a settler-colonial country, and the current and historical colonisation of Māori has served to dispossess Māori from their land, destroy their economic base, and threaten their culture and language (Lawson-Te Aho et al., 2019). This has led to them experiencing homelessness at disproportionately high rates, which is reflected in the following data used in our analysis.

2. Methods

This paper is a continuation of earlier analysis on the service usage in the HF cohort that utilised administrative and service-based records linked in the Integrated Data Infrastructure (IDI). This allows for linking of de-identified data for the 390 individuals in this group across a wide range of government interactions (Black, 2016). More detail on the IDI, including the datasets used, can be found in this original paper (Pierse et al., 2019). See Section 6 for the Statistics New Zealand (SNZ) disclaimer.

2.1. Study and comparison populations

The results are presented separately for the 210 women and 180 men in the HF cohort. This paper has used the same control group representative of the general population (n = 33,666) as in the initial analysis (hereafter referred to as the Estimated Resident Population, or ‘ERP’) (Gibb, Bycroft, & Matheson-Dunning, 2016; Pierse et al., 2019). There were 16,884 women in the ERP (50.2%).

2.2. Datasets

The demographic information presented in Table 1 was obtained using an IDI composite table with the most reliable estimate of that person's sex, age, and ethnicity. Information on parenthood was sourced from a government-maintained set of life events that links people to individual children's birth certificates which list them as a parent (Statistics New Zealand, 2015).

Table 1.

Demographics of the men and women in the HF and ERP cohorts.


Variable		 Relative percentage (%)
Housing First (n = 390)
 Relative percentage (%) ERP (n = 33,666)
Women (n = 210) Men (n = 180) Women (n = 16,884) Men (n = 16,785)
Age (years) Under 25 18.6% 13.3% 13.8% 15.8%
25–44 57.1% 45% 35.6% 36.5%
45–64 24.3% 41.7% 35.3% 33.6%
65+ S1 S 15.4% 14.1%
Ethnicity (total response, multiple ethnicities allowed) Māori 78.6% 66.7% 14.4% 14.1%
European 32.9% 48.3% 71.5% 69.3%
Pacific 7.1% 6.7% 6.5% 6.6%
Asian 2.9% S 13.8% 14.2%
Middle Eastern, Latin
American, African		 4.3% 6.7% 2.2% 2.7%
Other S S 1.5% 1.9%
Number of children, as listed on child's birth certificate None 18.6% 38.3% 55.9% 57.1%
1 17.1% 21.7% 16.2% 16.9%
2 21.4% 16.7% 16.4% 16.7%
3 15.7% 8.3% 7.2% 6.6%
4+ 27.1% 15% 4.3% 3.7%
Open in a new tab
1

Any count of an associated statistic with an underlying count of people or events below six (or 20 for a mean) are suppressed by SNZ for privacy and confidentiality reasons (as indicted by S in the tables).

The rates and types of service usage by sex are presented for datasets that have been grouped into one of three domains:

  • ●

    Health includes all publicly funded hospital discharges, subsidised pharmaceutical dispensings, and outpatient events from secondary mental health and addiction services.

  • ●

    Justice includes alleged criminal offences, all laid criminal charges, all convicted criminal court charges, and all correctional events.

  • ●

    Social development and tax information is sourced from two income datasets recorded monthly: wages and salaries, and main working-age welfare entitlements. Information on benefit type is sourced from administrative records about new benefit spells.

2.3. Reference period

The findings presented by sex are for events prior to the date first housed for HF individuals (between October 2014 and June 2017) and the median date at which the HF cohort were first housed (June 9, 2016) for the ERP (the ERP were not in any housing programmes, hence why we use the HF median date for their results). Results are presented for the five year and one year periods prior to the relevant baseline. That is, baseline here with respect to the HF individuals means the point at which those in the cohort were housed and the June 9, 2016 for the ERP.

2.4. Analysis

For each cohort of interest, the results are presented as exact numbers, means and, where appropriate, relative percentages. All analysis was done on de-identified records in a secure Data Lab environment and the necessary privacy, confidentiality, and security measures for IDI research have been observed. The ERP data comes from the 20181020 refresh of the IDI and the HF data from the 20190420 refresh; the HF data is from a later refresh due to data coverage quality and lags in updates.

3. Results

The results are presented below, first by comparing the key demographics of the HF women and men. We then move on to explore rates of service usage by different dataset domains, comparing the HF women to the HF men and to the ERP women. Finally, we look at the types of welfare recipiency and compare the difference between the HF women and men.

There were 390 people in the HF cohort, 210 (53.8%) of whom were women. Chi-square tests2 showed the women were more likely to be young (p < 0.001), Māori (p < 0.008), and have one or more children (p < 0.001) in comparison to the men. The most common age bracket for women in the cohort was 25–44, with 45–64 the next most common age bracket. However, there was a higher percentage (57.1%) of women aged 25–44, in comparison to the men (45%).

As Table 1 shows, 78.6% of HF women were Māori, which is higher than the 73.1% of the entire HF cohort who identified as Māori. HF women had more children (as listed on a birth certificate) than men: 81.4% of women had children in comparison to 61.7% of men.3 HF women were more likely to have four or more children (27.1%), whereas men were more likely to have only one child (21.7%).

Overall, women in the HF cohort had more children than the men. HF women had an average of 2.6 children, with a quarter of them having four or more children. In comparison, HF men had an average of 1.6 children, and a quarter of them had two or more children.

The HF cohort had significantly higher rates of service usage than the ERP. We compared women and men within the HF cohort, and HF women to ERP women; Table 2 shows these results for both five years and one year before baseline. We present both numbers to show the prolonged high rates of service usage that the cohort experienced prior to receiving support from TPP. Overall, within the HF cohort, women and men had comparable rates of healthcare usage; men had higher justice interactions; and women earned significantly less from wages and salaries. Additionally, when comparing the HF women to ERP women, we see that the HF women had much higher rates of service usage across all domains, and considerably higher welfare recipiency and lower wages.

Table 2.

Comparative rates of service usage.

Dataset domain Data source Mean in 5 years before baseline
 Mean in 1 year before baseline
Women
 Men
 Women
 Men
HF ERP HF ERP HF ERP HF ERP
Health Hospitalisations 3.8 1.1 2.5 0.8 0.8 0.2 0.7 0.2
Maternity-related hospitalisations1 1.3 0.3 S S 0.2 0.1 S S
Pharmaceutical dispensings 128.0 64.9 157.5 53.2 32.2 14.9 31.9 12.5
Mental Health & Addiction - Outpatient events2 69.0 5.8 76.6 5.9 16.1 1.3 21.0 1.2
Justice Police offences 2.2 0.1 5.6 0.5 0.4 <0.1 1.2 0.1
Criminal charges, laid 1.8 0.1 5.5 0.5 0.3 <0.1 1.1 0.1
Criminal charges, convicted 1.2 0.1 4.0 0.3 0.2 <0.1 0.8 <0.1
Prison sentence, days 22.3 0.4 213.3 10.1 0.77 0.04 41.7 1.6
Community Service sentence, days 95.6 5.0 214.8 26.7 28.3 0.9 44.6 10.5
Social Development and Tax Months in which tax paid on wages and salaries 8.4 29.1 10.1 30.6 1.3 6.3 1.4 6.7
Total income from wages and salaries $18,886 $105,681 $23,953 $161,961 $2279 $23,999 $2645 $36,802
Months in which a benefit was received 44.7 7.1 38.9 5.1 10.2 1.4 8.9 1.0
Total income from benefit receipt $56,829 $8222 $41,135 $5195 $13,557 $1692 $10,041 $1131
New benefit spells 2.1 0.4 2.9 0.3 0.6 0.1 0.8 0.1
Open in a new tab
1

All hospitalisations with primary diagnosis classification of ‘Complications of pregnancy, childbirth, and the puerperium’ or supplemental classifications V20–V39 (‘Persons encountering health services in circumstances relating to reproduction and development,’ ‘live-born infants according to type of birth’).

2

Outpatient rates of mental health and addiction service usage have been found from a single source of national-level data about contacts, activities, and services for secondary-care mental health and addiction service providers.

Healthcare service usage was the first domain explored. Overall, there was not a significant difference between the healthcare usage of women and men; however, both had significantly higher usage than their ERP counterparts. For example, in the five years prior to baseline, the mean number of hospitalisations was 3.8 for HF women versus 1.1 for ERP women, and 2.5 for HF men versus 0.8 for ERP men. HF women were 3.5 times more likely to be hospitalised than ERP women. Both HF women and men had higher rates of healthcare usage than the ERP. However, in comparison to women in the ERP, HF women had distinct healthcare usage. For HF women, for example, maternal hospitalisations in the five years prior to being housed had a mean of 1.3, whereas during the same period the ERP had a mean of 0.3; HF women were 4.3 times more likely to experience maternal hospitalisations. Furthermore, if we look at the hospitalisation rate per birth, we find that the HF women have higher maternal hospitalisations per birth than the ERP women. HF women had an average of 1.17 hospitalisations per birth, while ERP women had an average of 0.83 hospitalisations per birth. This suggests that the HF women have distinct, more acute, healthcare needs, particularly in relation to maternity care; they are not simply seeing more maternal hospitalisations than the ERP because they have a higher number of children—each birth, on average, sees them hospitalised at a higher rate than for ERP women. While HF women had distinct healthcare usage, their overall usage was not significantly different to HF men, it did, however, differ to the ERP women. These trends remain similar in the year prior to becoming housed.

The second domain explored was justice. As Table 2 shows, HF women had significantly fewer interactions across justice datasets than the men, although they had higher justice interactions than ERP women. HF women spent a vastly smaller number of days in prison (22.3) than the men (213.3). This difference was slightly less pronounced for days spent doing community service. Additionally, HF men were 2.5 times more likely to have a recorded police offence than HF women.

The third domain explored was social development and tax. HF women had significantly less time in paid employment than ERP women and HF men; they were more likely to be receiving income from a benefit. Both women and men in the HF cohort earnt significantly less from wages and salaries than their counterparts in the ERP, and had higher rates of benefit receipt than the ERP. In the five years before baseline, HF women's total income from wages, salaries, and benefits was a mean of $75,715 compared to a mean of $113,903 for ERP women. Women spent more months receiving a benefit than men, with a mean of 44.7 months in comparison to 38.9 months in the five years before baseline. We also see in the year prior to becoming housed that the HF cohort—particularly the women—had incomes far lower than the ERP.

Data on the distinct types of benefits that the HF cohort had received as primary4 recipient, based on administrative records of new spells commenced, is displayed in Table 3. For this analysis, we grouped benefit receipt into 11 types of benefits capturing similar purpose entitlements over time. Table 3 displays the results for the five most common benefit types. There were stark gender differences in our HF data; for women, the most common type of benefit was Sole Parent support—which indicates the women had dependent children living with them—and for men it was the Jobseeker benefit.5

Table 3.

Types of benefit receipt and counts of weeks and people receiving entitlements, by five most common (in weeks) benefit types in the HF cohort in the five years and one year before housed.

Type of benefit 5 years before housed
 1 year before housed
Women (n = 210)
 Men (n = 180)
 Women (n = 210)
 Men (n = 180)
Number of people in receipt1 Average number of weeks per person (n = 210) Number of people in receipt Average number of weeks per person (n = 180) Number of people in receipt Average number of weeks per person (n = 210) Number of people in receipt Average number of weeks per person (n = 180)
Sole Parent 123 (58.6%) 85 12 (6.7%) 6 93 (44.3%) 17 9 (5%) 2
Invalids 45 (21.4%) 38 54 (30%) 47 39 (18.6%) 8 48 (26.7%) 12
Sickness 75 (35.7%) 22 14 (63.3%) 56 41 (20%) 6 69 (38.3%) 12
Jobseeker 81 (38.6%) 25 111 (61.7%) 45 51 (24.3%) 8 81 (45%) 11
Caring Sick Infirm 9 (24.3%) 2 S S S S S S
Open in a new tab
1

These equate to more than 100% of the cohort as individuals can be on more than one benefit at a time.

Many more HF women than men were sole parents in need of financial support. In the five years prior to baseline, there were 123 women receiving a Sole Parent benefit, compared to only 12 men. This meant 58.6% of the HF women were receiving a Sole Parent benefit, suggesting large numbers of women with children were living in poverty in the five years prior to being housed. As will be raised in the Discussion (section 4), benefit rates in Aotearoa New Zealand are incredibly low. Gender norms mean that overall, women do more child-rearing, and experience more poverty (Statistics New Zealand, 2013, 2014a). Welfare states were created, in part, to support women and children in instances where husbands were unable to support their families (Orloff, 1996). However, despite changing gender norms such as women's increased participation in the labour force, and it being somewhat more socially acceptable to raise a child as a single mother, the neoliberalisation of the welfare state has demonstrated the continued vulnerability of women to gendered systems that devalue domestic labour. The data presented shows the starkest end of this dynamic, in which women's experiences of homelessness differs to that of men's; in particular, that they are much more likely to be reliant on government support for sole parents. Benefits rates must be increased to ensure that people are not trapped in poverty, and gender norms need to continue to be challenged to better support an equal division, and valuing, of domestic labour.

4. Discussion

Homeless women in Aotearoa New Zealand experience a number of hardships and frequently find themselves receiving inadequate support from the welfare system. The data presented builds on previous research from He Kāinga Oranga by showing these women have a higher rate of interactions with government agencies than the ERP women in the years leading up to them needing housing assistance from TPP (Pierse et al., 2019). We present data from both the five- and one-year periods prior to being housed in order to highlight that the needs of this group do not suddenly occur, and that they are not necessarily “hard to reach.” If government systems were functioning as intended, our cohort should not have required assistance from TPP. In particular, the welfare system is clearly not providing adequate income support, as we see that the women in this cohort had very small incomes in the years leading up to their engagement with TPP. There are repeated points in the entire five years prior to needing support at which vulnerable women present with needs that are not adequately met, resulting in them becoming homeless and needing the support of TPP. In particular, many women with children are living in poverty and ultimately become homeless.

The primary limitation of this paper is that while the administrative data used allows for many unique and interesting analyses, there is likely to be an undercount of the service usage of the cohort. This is due to the likely possibility of some data being missing due to a lack of records; the mental health data, for example vary greatly in how they are reported by individual agencies. Additionally, the analyses presented may not be applicable to the entire homeless population in Aotearoa New Zealand, as the HF cohort is relatively small. However, the demographics of our cohort are similar to that of the broader population of people who experience homelessness in Aotearoa New Zealand; they are roughly 50% women, younger, and more likely to be Māori (Amore et al., 2020).

As mentioned above, the wider homeless population in Aotearoa New Zealand, which includes people living in overcrowded and emergency housing, sees a roughly 50/50 split between men and women (Amore et al., 2020).6 This differs from the international literature in which homelessness—particularly rough sleeping—is generally presented as an issue primarily affecting men (Bretherton, 2017; North & Smith, 1993; Phipps et al., 2018; Pleace, 2016; Reeve, 2018; Velasquez & Larose, 2015). Within the existing homelessness literature, domestic violence, which is more often experienced by women, is frequently found to be a precursor to homelessness (Hagen, 1987; Hagen & Ivanoff, 1988; May et al., 2007; Tessler, Rosenheck, & Gamache, 2001; Wardhaugh, 1999). For this paper, we have not been able to contribute to this body of evidence due to limitations in identifying domestic violence from the available administrative and service-based data in the IDI. Motherhood has an enormous influence on the circumstances in which women become homeless. The existing literature reports mothers were more likely to be homeless than women without children, but were homeless for shorter periods of time (Johnson & Kreuger, 1989). As shown throughout this paper, 81.4% of the HF women were mothers, many of whom had multiple children, and were receiving the inadequate Sole Parent benefit (Whakamana Tāngata, 2019).7 One positive aspect of homeless women's lives when viewing their service usage in other studies is that they have lower rates of incarceration and felony convictions than homeless men (Calsyn & Morse, 1990; North & Smith, 1993). As discussed, this data showed similar findings; the HF women had fewer interactions across justice datasets than HF men and spent far less time in prison than HF men.

Māori, particularly Māori women, were over-represented in the HF cohort. As a whole, Māori women are at a high risk of poverty and discrimination (Statistics New Zealand, 2014b). In the context of homelessness, Māori women face significant discrimination in the housing market, and high rates of imprisonment, which are both drivers of homelessness (Cormack, Harris, & Stanley, 2019; Smale, 2020). This, alongside the data presented showing the disproportionate number of Māori in our cohort, indicates a need for Māori-centred and Māori-led support that accounts for the cultural aspect of homelessness in Aotearoa New Zealand, with a particular focus on Māori women (Lawson-Te Aho et al., 2019).

Most HF women were aged 25–44, which is the age range at which women are most likely to be having, or caring for, children. This aligns with the data that indicated most of the HF women (81.4%) were parents. It is not possible to tell how many of these women had children in their custody at the time of being housed by TPP, but international literature indicates that homeless women are more likely to have their children in their custody than homeless men (North & Smith, 1993). Supporting this, the benefit data showed that the majority of HF women who were receiving a benefit in the years before being housed were on Sole Parent benefits, indicating that a large number of them had children in their custody.

In liberal welfare states (such as Aotearoa New Zealand, the United Kingdom, and Australia), neoliberalism and consequent decreases in state support have resulted in high levels of poverty (Stephens & Fitzpatrick, 2007). In social democratic welfare states (such as the Nordic countries) that have maintained high levels of welfare support, poverty has remained relatively low (Stephens & Fitzpatrick, 2007). Liberal welfare states see fewer women in full time employment, alongside high childcare costs, whereas social democratic welfare states see more women in full time employment (Stephens & Fitzpatrick, 2007). Social democratic welfare states are also more likely to provide support for single mothers than liberal welfare states (Bretherton, Benjaminsen, & Pleace, 2017). These robust welfare systems are protective factors in preventing homelessness. Aotearoa New Zealand can be categorised as a liberal welfare state, which as per Stephens and Fitzpatrick (2007) means poverty, income inequality, and homelessness are likely to be high. As has been shown, our results reflect this.

Benefit and income data show that these women were low-income earners in need of government support. A 2019 government mandated Welfare Expert Advisory Group (WEAG) review of the welfare state in Aotearoa New Zealand reported that the “current welfare system is no longer fit for purpose and needs fundamental change” (Whakamana Tāngata, 2019, p. 5). Amongst the 42 recommendations the WEAG gave was an urgent request to raise main benefit rates by up to 47%; stating “current levels of support fail to cover even basic costs for many people, let alone allowing them to meaningfully participate in their communities” (Whakamana Tāngata, 2019, p. 7). The current structuring of the welfare state in Aotearoa New Zealand does not allow for beneficiaries to live lives as dignified or respected as their non-beneficiary counterparts. Benefits in Aotearoa New Zealand are low, and there is a strong association between receiving a benefit and living in poverty (Whakamana Tāngata, 2019). The WEAG's report noted Aotearoa New Zealand has a high rate of sole parenthood, sole parent benefit receipt, and a high rate of poverty amongst sole parent families (Whakamana Tāngata, 2019). Additionally, the rate of sole parent benefits does not change if a person has multiple children, which further pushes women into poverty. At present, aspects of the welfare system in Aotearoa New Zealand do not support women's role as carers, despite women's high rate of benefit receipt (Whakamana Tāngata, 2019). Additionally, a greater number of Māori women receive benefits than non-Māori, and Māori men (Whakamana Tāngata, 2019). Thus, lifting benefit rates will help to lift children and families out of poverty (Whakamana Tāngata, 2019). As discussed earlier, the neoliberalisation of the welfare state has demonstrated the continued vulnerability of women to gendered systems that devalue domestic labour. Lifting benefit rates is one way in which this can begin to be addressed.

5. Conclusion

Women's homelessness is a pressing issue both globally and in Aotearoa New Zealand, yet there remains limited scholarship about the phenomenon. This paper contributes to the knowledge base by exploring the service usage of a cohort of formerly homeless people in an Aotearoa New Zealand city. The IDI provides us with a unique ability to be able to look at interactions with government services for various subsets of Aotearoa New Zealand's population. These data have been used to explore the gender differences of homelessness amongst a Housing First cohort and a comparison with women in a control group representative of the general population. The wider HF cohort were primarily women, young, Māori, and parents. This paper has focused specifically on the women in this cohort in order to explore the gendered nature of homelessness. They faced significant poverty, earning much less than their non-homeless counterparts, and relied heavily on government support. The data presented shows that the neoliberalisation of the welfare state, high rates of women's poverty, and the gendered dynamics of parenthood are factors contributing to women's homelessness being distinct to men's homelessness.

6. Statistics New Zealand disclaimer

These results are not official statistics. They have been created for research purposes from the Integrated Data Infrastructure (IDI) which is carefully managed by Stats NZ. For more information about the IDI please visit https://www.stats.govt.nz/integrated-data/.

The results are based in part on tax data supplied by Inland Revenue to Stats NZ under the Tax Administration Act 1994 for statistical purposes. Any discussion of data limitations or weaknesses is in the context of using the IDI for statistical purposes, and is not related to the data's ability to support Inland Revenue's core operational requirements.

Funding

This paper was supported by funding from the New Zealand Ministry of Business Innovation and Employment, Endeavour Fund.

Ethics statement

Ethics Ethical approval was given by the University of Otago Human Research Ethics Committee ref HD16/049.

Acknowledgements

We acknowledge the 390 individuals that are discussed in this paper, each of whom has their own unique story to tell. We also acknowledge our research partners at The People's Project and the University of Waikato, without whom this research would not be possible.

Footnotes

1

We use the term women in this paper for the reader's ease, however, due to most government data sources only collecting data about sex assigned at birth, the analysis is of those who were assigned female at birth; 90% of the women in the cohort had their sex identified by Births, Death, and Marriages registration data and the remaining 10% from alternative government datasets. We were unable to tell how many of these people identified as women. Thus, it may be that there are some non-binary people or transgender men included in the analysis without our knowledge. Additionally, the analyses will not include any transgender women.

2

Chi-square tests are a statistical method used to determine whether or not there are statistically significant differences between variables.

3

In Aotearoa New Zealand a parent, i.e. particularly a father, is able to be left off the birth certificate.

4

Primary recipient here either means a sole recipient, or the main recipient when a partner has been declared.

5

Sole Parent support does not increase with the number of children a family has.

6

Internationally, it is rare for homeless populations to see such a high proportion of women. One of the main reasons for Aotearoa New Zealand seeing such a high proportion of women experiencing homelessnesss is due to our comprehensive definition of homelessness, which picks up on so-called “hidden” homeless populations. The definition is a national one utilised by SNZ, and counts are conducted during censuses. For more about how homelessness is defined and measured in Aotearoa New Zealand, refer to the work of Dr. Kate Amore. We also believe that other factors such as colonisation and an inadequate welfare system contribute to this, however, we also know that these are not unique to Aotearoa New Zealand. More research is needed to investigate this.

7

Our research partners at TPP note that they are not sure why the cohort is primarily women, but that they think it is in part due to one of their first successful clients being a woman. TPP believe she then told other women about their services, leading to an increase in women accessing their services. However, we do not necessarily feel that women with dependent children were more likely to be accepted in the programme. This is because while TPP initially accepted anyone who asked for support, they quickly became overwhelemed and began using the VI-SPDAT and their own knowledge of the local context to triage clients, and to help them to deal with funding requirements and resourcing constraints. This meant that they were required to direct families to other agencies and teams focused solely on supporting families experiencing homelessness.

References

  1. Amore K., Viggers H., Howden-Chapman P. 2020. Severe housing deprivation in Aotearoa New Zealand, 2018. Wellington. [Google Scholar]
  2. Black A. 2016. The IDI prototype spine's creation and coverage. (Statistics New Zealand Working Paper No 16-03)www.stats.govt.nz Wellington. [Google Scholar]
  3. Bretherton J. Reconsidering gender in homelessness. European Journal of Homelessness. 2017;11(1) http://www.feantsaresearch.org/download/feantsa-ejh-11-1_a1-v045913941269604492255.pdf [Google Scholar]
  4. Bretherton J., Benjaminsen L., Pleace N. Women's homelessness and welfare states. In: Mayock P., Bretherton J., editors. Women's homelessness in europe. 2017. pp. 75–102. [DOI] [Google Scholar]
  5. Bukowski K., Buetow S. Making the invisible visible: A photovoice exploration of homeless women's health and lives in central auckland. Social Science & Medicine. 2011;72(5):739–746. doi: 10.1016/j.socscimed.2010.11.029. [DOI] [PubMed] [Google Scholar]
  6. Calsyn R.J., Morse G. Homeless men and women: Commonalities and a service gender gap. American Journal of Community Psychology. 1990;4:597–608. doi: 10.1007/BF00938062. August. [DOI] [PubMed] [Google Scholar]
  7. Cormack D., Harris R., Stanley J. 2019. Māori experiences of multiple forms of discrimination: Findings from Te kupenga 2013. [DOI] [Google Scholar]
  8. Gibb S., Bycroft C., Matheson-Dunning N. 2016. Identifying the New Zealand resident population in the integrated data infrastructure (IDI)http://archive.stats.govt.nz/methods/research-papers/topss/identifying-nz-resident-pop-in-idi.aspx Wellington. [Google Scholar]
  9. Groot S., Hodgetts D., Waimarea Nikora L., Leggat-Cook C. A Māori homeless woman. Ethnography. 2011;12(3):375–397. doi: 10.1177/1466138110393794. [DOI] [Google Scholar]
  10. Hagen J.L. Gender and homelessness. Social work. 1987:312–316. https://academic.oup.com/sw/article-abstract/32/4/312/1888688 (July-August) [Google Scholar]
  11. Hagen J.L., Ivanoff A.M. Homeless women: A high-risk population. Affilia: Journal of Women and Social Work. 1988;3(1):19–33. https://journals.sagepub.com/doi/pdf/10.1177/088610998800300103 [Google Scholar]
  12. Johnson A.K., Kreuger L.W. Toward a better understanding of homeless women. Social Work. 1989;34(6):537–540. https://www.jstor.org/stable/pdf/23715703.pdf?refreqid=excelsior%3A30450560e5697bdd51d530c45d28cc26 [PubMed] [Google Scholar]
  13. Lawson-Te Aho K., Fariu-ariki P.F., Ombler J., Aspinall C., Howden-Chapman P., Pierse N. A principles framework for taking action on Māori/Indigenous Homelessness in Aotearoa/New Zealand. SSM - Population Health. 2019;8 doi: 10.1016/j.ssmph.2019.100450. July. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. May J., Cloke P., Johnsen S. Alternative cartographies of homelessness: Rendering visible British women's experiencing of “visible” homelessness. Gender, Place & Culture. 2007;14(2):121–140. doi: 10.1080/09663690701213677. [DOI] [Google Scholar]
  15. North C.S., Smith E.M. A comparison of homeless men and women: Different populations, different needs. Community Mental Health Journal. 1993;29 doi: 10.1007/BF00754410. https://link.springer.com/content/pdf/10.1007%2FBF00754410.pdf [DOI] [PubMed] [Google Scholar]
  16. Orloff A. Gender in the welfare state. Annual Review of Sociology. 1996;22(1):51–78. doi: 10.1146/annurev.soc.22.1.51. [DOI] [Google Scholar]
  17. Phipps M., Dalton L., Maxwell H., Cleary M. Women and homelessness, a complex multidimensional issue: Findings from a scoping review. Journal of Social Distress and the Homeless. 2018;28(1):1–13. doi: 10.1080/10530789.2018.1534427. [DOI] [Google Scholar]
  18. Pierse N., Ombler J., White M., Aspinall C., McMinn C., Atatoa-Carr P. Service usage by a New Zealand Housing First cohort prior to being housed. SSM - Population Health. 2019;8 doi: 10.1016/j.ssmph.2019.100432. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Pleace N. Exclusion by definition: The under-representation of women in European homelessness statistics. In: Mayock P., Bretherton J., editors. Women's homelessness in europe. Palgrave Macmillan; London: 2016. pp. 105–126. [DOI] [Google Scholar]
  20. Reeve K. Women and homelessness: Putting gender back on the agenda. People, Place and Policy. 2018;11(3):165–174. doi: 10.3351/ppp.2017.8845235448. [DOI] [Google Scholar]
  21. Smale A. Rough justice. Radio New Zealand. 2020. https://shorthand.radionz.co.nz/maori-justice-racism/index.html
  22. Statistics New Zealand . 2013. Caring for Children : Findings from the 2009/10 time use survey. Wellington. [Google Scholar]
  23. Statistics New Zealand . 2014. 2013 census QuickStats about income.http://archive.stats.govt.nz/Census/2013-census/profile-and-summary-reports/quickstats-income/personal-income-qualification.aspx%0Ahttp://archive.stats.govt.nz/Census/2013-census/profile-and-summary-reports/quickstats-income/personal-income-ethnic.aspx Wellington. [Google Scholar]
  24. Statistics New Zealand . 2014. 2013 QuickStats about Māori. Wellington. [Google Scholar]
  25. Statistics New Zealand . 2015. IDI data dictionary: Life event data (july 2015 edition)http://archive.stats.govt.nz/browse_for_stats/snapshots-of-nz/integrated-data-infrastructure/idi-data/life-event-data.aspx Wellington. [Google Scholar]
  26. Stephens M., Fitzpatrick S. Welfare regimes, housing systems and homelessness: How are they linked? European Journal of Homelessness. 2007;1:201–212. December. [Google Scholar]
  27. Tessler R., Rosenheck R., Gamache G. Gender differences in self-reported reasons for homelessness. Journal of Social Distress and the Homeless. 2001;10 https://link.springer.com/content/pdf/10.1023%2FA%3A1016688707698.pdf [Google Scholar]
  28. Tsemberis S. Housing first: Ending homelessness, promoting recovery, and reducing costs. In: Ingrid G.E., O'Flaherty B., editors. How to house the homeless. Russell Sage Foundation; New York: 2010. pp. 37–57. [Google Scholar]
  29. Whakamana Tāngata Wellington. 2019. www.weag.govt.nz
  30. Velasquez A., Larose R. Youth collective activism through social media. The role of collective efficacy. 2015;17(6):899–918. doi: 10.1177/1461444813518391. [DOI] [Google Scholar]
  31. Wardhaugh J. The unaccommodated woman: Home, homelessness and identity. Sociological Review. 1999;47(1) doi: 10.1111/1467-954X.00164. [DOI] [Google Scholar]

Articles from SSM - Population Health are provided here courtesy of Elsevier

RESOURCES