WANGANUI AFFORDABILITY STUDY The potential impact on households physical and mental health of rates increases to fund upgrading Wanganui s wastewater scheme A report commissioned by Wanganui District Council Department of Public Health The Wellington School of Medicine May 2000
Contributors Des O Dea Health Economist, Department of Public Health, Wellington School of Medicine Anna Matheson Research Assistant, Department of Public Health, Wellington School of Medicine Advisory Group Dr Philippa Howden-Chapman Senior Lecturer in Health and Public Policy, Department of Public Health, Wellington School of Medicine Dr Tony Blakely Public Health Specialist, Department of Public Health, Wellington School of Medicine Dr Peter Crampton Senior Lecturer in Health Management, Department of Public Health, Wellington School of Medicine Acknowledgements Craig Roberts (Wanganui District Council)
CONTENTS CONTENTS...1 TABLES AND FIGURES...2 TABLES...2 FIGURES...2 SUMMARY...3 1. CONTEXT...5 2. METHODOLOGY...6 2.1 OBJECTIVES...6 2.2 RESEARCH APPROACH...6 2.3 CONCEPTUAL MODEL...7 3. DETERMINANTS OF HEALTH...8 4. INCOME...10 4.1 WANGANUI DISTRICT AND ALL NEW ZEALAND COMPARED...10 4.2 RELATIVE POVERTY AND EQUIVALISED INCOME...11 4.3 GROUPS BELOW THE POVERTY LINE...13 5. IMPACT ON HOUSEHOLD BUDGET...15 5.1 HOUSEHOLD EXPENDITURE PATTERNS IN NEW ZEALAND...15 5.2 HOUSEHOLD EXPENDITURE PATTERNS IN WANGANUI DISTRICT...19 6. HOUSEHOLDS AT RISK OF BEING AFFECTED BY A RATES INCREASE...19 6.1 THE EFFECT OF RELATIVE POVERTY ON HOUSEHOLD CONSUMPTION...20 6.2 THE EFFECT OF A CUT IN INCOME ON HOUSEHOLD SPENDING PATTERNS....21 7. AT-RISK HOUSEHOLDS AND THEIR HEALTH...22 7.1 HEATING COSTS...24 7.2 FOOD SECURITY...24 8. THE ACCOMMODATION SUPPLEMENT...24 9. CONCLUSIONS...28 REFERENCES...29 APPENDIX 1 TERMS OF REFERENCE...30 APPENDIX 2...34 1
TABLES AND FIGURES Tables 1 Mean and Median 1996 Census Household Incomes: Wanganui and New Zealand. 2 Mean and Median Equivalised 1996 Census Household Incomes: Wanganui and New Zealand. 3 Proportions of households and persons below the poverty line, Wanganui and New Zealand. 1996 census. 4 Characteristics of individuals living in Wanganui households below the 1996 census poverty line. 5 Household expenditure by category all New Zealand. 1997/98 6 Comparison of Housing and Total Expenditure; Wanganui District and New Zealand. 7 Effect of reduction in Overall Household Budget from increase in housing costs. 8 The Accommodation Supplement. Income limits and accommodation cost thresholds. Figures 1 Comparison of distribution of 1996 census household incomes. Wanganui District and New Zealand. 2 Housing (adjusted) expenditure as percent of total household expenditure. All NZ households 1997/98. By census income groups. 3 Expenditure breakdown by household census income groups. All New Zealand. 1997/98. 4 Housing (adjusted) expenditure as percent of total household expenditure. All NZ households 1997/98. By household type. 5 Housing (adjusted) expenditure as percent of total household expenditure. All NZ households 1997/98. By Tenure type. 6 Adult Females aged 15-64: State of Health Excellent or Very Good by Income Quintiles. 7 Adult Males aged 15-64: State of Health Excellent or Very Good by Income Quintiles. 8 Wanganui City Area Units Average Rents against Deprivation Index. (Excluding rural area units.) 2
SUMMARY Inequalities in the basic distribution of material resources, income, education, employment and housing, generate health inequalities. Research has found that within New Zealand there is a strong and consistent relationship between socioeconomic factors and health outcomes such as poorer self-perceived health, hospitalisation, morbidity and premature mortality. It has been found that people who are better educated, have professional occupations, higher incomes and live in less socioeconomically deprived neighbourhoods are highly likely to enjoy better health and live longer than those who have no qualifications, are unemployed, earn less and live is socioeconomically deprived neighbourhoods. All sectors of society are affected: there are no neat cut-off points. As a result, health status is distributed as a gradient up the social hierarchy and applies to almost all causes of death. Not only do these factors powerfully predict health and premature mortality at a given point in time, but these health inequalities persist over generations. At the time of the 1996 census, the mean income for Wanganui households was $34,705 compared with $45,926 for All New Zealand. The equivalised income (which takes into account the number of people living in a household and its composition), for Wanganui was $31,980 compared with $40,853 for All New Zealand. If an ad hoc poverty line is drawn at an income of $15,000 it is shown that 16.7% of households in Wanganui live below the poverty line compared with 12.9% of households of persons for All New Zealand. One in six of all households in Wanganui are below this poverty line. Maori are disproportionately represented accounting for a third of those below the poverty line. As Maori non-response rates to the census household income question this figure is likely to be even higher. Nearly a quarter of all children in Wanganui live below this poverty line. Relative poverty is most prevalent in low-income families with dependent children, including sole parent families. According to a study on the effect of relative poverty on household consumption, many low income New Zealand households are substantially deprived of essential household items and services. 3
Housing costs vary more with the life-stage of a household, that is the age of the adults and whether or not they have dependent children, than they do with household income. From a recent study it has been found that families below the poverty line do sometimes forego medical treatment they feel is needed, and sometimes cannot make purchases or pay bills for items important to family health. There is New Zealand evidence that low-income families are more food insecure, that is to economise, they can only afford to eat properly sometimes and that due to lack of money they sometimes run out of food and eat less because of lack of money. There is also evidence that older people in particular economise on fuel bills and that this may be one of the reasons New Zealand has greater seasonal mortality among those 65 years and older that in other countries with worse climates. Conclusions Households most likely to be at risk from a change in discretionary income affecting health and wellbeing are low-income families with children, both sole parent and couple families, either renting or paying off a mortgage. The majority of persons in such households are non-maori, but a relatively high proportion of all Maori live in such households. A quarter of those living under the poverty line are children. The physical health of such families could be at risk by insufficient spending on nutritious food, on heating, and on medical care. A recent Family Centre report on low income families shows that many have difficulty in keeping up essential outlays on all these items. Psychological health is more difficult to relate to consumption choices, but poor families do have difficulty in participating in many social activities. For the average household, any reduction in household discretionary income is likely to be across all categories of spending but less so for food and household operations. To some extent the Accommodation Supplement will help offset the effect of a rates or rent increase on low income households. If they qualify, and have unused entitlement to the Accommodation Supplement, this could be used to meet 70% of the cost of any rates or rent increase. Note that not all qualifying households actually take up the supplement and we are not aware of any estimates of uptake that could indicate the extent to which those entitled to the supplement do not know they are entitled or do not how to get it. These numbers could be significant. 4
1. CONTEXT The current wastewater scheme is a significant financial burden on Wanganui ratepayers. Major environmental gains have already been achieved since the project s inception. Sewerage no longer enters the Wanganui River, except during some flood events. Council is working on modifying the wastewater scheme so that the additional environmental benefits for the community should become affordable by ratepayers for the community. The expenditure in Council s 10 year Plan reflects a scenario that involves the treatment of effluent in aerated lagoons, followed by clarification, then ultraviolet disinfection and disposal through the sea outfall. The project also involves separation of stormwater drains from sewers. The wastewater scheme review is now well under way, with improved and more affordable scheme details having been developed. The application for the review is scheduled to be submitted to horizons.mw (Regional Council) in October 2000. The Wanganui District Council is commissioning the research and analysis of a series of questions in relation to the impact on individuals and the community of increasing rates, to pay for the upgrade of Wanganui s wastewater scheme. The Council has commissioned the Department of Public Health at the Wellington School of Medicine to attempt to answer these three questions. The expected rates increase required to fund the upgrade is of the order of $100 on an average residential property. In addition to increased rates, private home owners will also incur separation capital costs associated with replacing one combined sewerage and stormwater pipe on their property with two (ie one for sewerage and one for stormwater). This private work will hook up into the new separated public wastewater / stormwater system. Newer subdivisions are already separated and residents in these will not face separation costs, however the majority of households will. Initial rough estimates are 8100 properties at $800 each and 1900 properties at $3000 each (ie costs will range from $800 to $3000). In annualised terms these capital outlays would approximate to a further $80 to $300 per year. Council has considered the option of providing loans to residents who need assistance with funding these capital outlays but as yet no decision on this has been made. 5
2. METHODOLOGY 2.1 Objectives To help Wanganui District Council answer the following three questions. Question 5: Question 6: Question 7: Is a decrease in household discretionary income likely to result in a decline of expenditure in those goods and services required to maintain the physical and mental health? What is the impact on household s financial capacity to maintain physical and mental health status if rates are increased in real terms by say $50 or $100 or $150? Under the existing rating system what is the Affordability Threshold for the Wanganui community before risk of significant harm eventuates? Note that work on Questions 1 to 4 (see Appendix 1) has been commissioned by Wanganui District Council from other consultants, Frances Sutton and BERL Ltd. Where appropriate this report draws on their material. It is not possible to promise precise answers to the above questions. Individual households will vary greatly in the way they respond to a cut in discretionary income. Most will have the resources to cope adequately, and will adjust their household spending in ways which do not harm the physical and mental health of household members. Some households, however, with limited resources, or perhaps using unwisely those resources they do have, will come under pressure. Our task, using existing data and research, is to give a general indication of the likely effects of a substantial increase in rates on households spending, and to indicate the kinds of household which will feel most pressure and whose physical and mental health might be affected. In what follows we have focused in particular on the relative importance of outlays on housing (rent, mortgage payments, rates, etc) for various kinds of household. This is based on the assumption that most households have little discretion over how much they spend on housing, so that changes in discretionary income are felt most in other consumption categories (food, apparel, transport, etc), and are felt more if housing costs already account for a large proportion of total household outlays. 2.2 Research Approach The research approach proposed for this report was to: (a) Tabulate average all New Zealand household spending patterns for selected household types from the Household Economic Survey (HES). (b) Make assumptions about the extent to which these might need adjusting to better fit Wanganui households for example lower housing costs. 6
(c) Using census material, make estimates of the number of low-income households of various types in Wanganui district. (d) Calculate, using the Michelini estimates, the likely consequences of reductions of the given magnitudes ($50, $100, $150) in discretionary income. (e) Cross-check with Family Centre results to the extent possible. (f) Tabulate also some recent Wellington School of Medicine research material on the relationship between household income and health. (g) Discuss and report on the implications of the material in (a) to (f), and also that on nutrition and heating. 2.3 Conceptual Model A household has a given budget B for all its annual purchases of goods and services. We assume that: (i) It allocates its spending in a way that is intended to maximise the well-being of household members. (ii) In doing this, specified minimums of certain commodities need to be purchased in order to assure the physical and psychological safety of those in the household. In particular physical safety requires the purchase of sufficient amounts of food, housing and health-care. Psychological well-being is aided by being able to participate in common social activities, and therefore by being able to make the outlays necessary for that purpose. (iii) A household s spending budget depends to a large extent on the total income coming into the household, though it can be adjusted to some extent by reducing or increasing savings, or reducing or increasing borrowing. (iv) The needs a household s spending budget must cover are to a large extent determined by the number of persons in the household, and also their age and whether or not they are in the paid work-force. Dwelling tenure is also important in determining what a household needs to spend its income on. Renting households and those paying off a mortgage in general have considerably higher outlays on housing than do households which own their dwelling mortgage-free. (v) An increase in the price of any one commodity reduces the amount available for spending on other goods and services. Quantities purchased of other goods and services will be reduced, except to the extent that levels of saving or borrowing are adjusted. (vi) The quantity purchased of the commodity whose price has increased will also normally reduce. (vii) In the case of housing, however, the assumption is made for this report that there will be zero fall in the quantity of housing purchased if the cost of housing goes up. That is, the extra spending on housing leads to less being purchased of 7
non-housing goods and services, but not less of housing. This assumption is not, it should be said, totally accurate. Some people might choose not to purchase the better house they would otherwise have bought, or some renters might double up. But most do not change the quantity or quality of their housing, particularly in the short-term. (viii) It is also assumed for this report that any increase in rates is passed on directly into increased rents on any rental property (for a differing point of view please see Appendix 2). This is because the total stock of available housing is more or less fixed at a given point in time, and also because every owner of rental property in the Wanganui urban area will face similar rates increases. Again this assumption will not be totally accurate (particularly with the proposed change in Housing NZ rentals, and if Housing NZ increases its stock of properties to satisfy a greater rental demand), but is a reasonable prediction of the likely outcome. Mathematically these assumptions might be expressed as: Budget B = Sum over all purchased commodities i of P I x Q I where P I is the price of the commodity and Q I the quantity purchased. B (P H xq H ) = Discretionary budget left after purchase of Q H of housing. For certain commodities we require Q I >= Q I, where Q I is the minimum required for physical and/or psychological safety. If P H increases, the discretionary budget reduces, and so therefore do quantities purchased Q I. In certain circumstances these might fall below Q I. In what follows we investigate these possibilities. It is possible, using recent research, to estimate the average amount by which households will reduce purchases, within broad expenditure groups. It is not possible to estimate precisely the levels of spending below which it is dangerous to fall. It is possible to give some indication, however, of those kinds of households which are at risk of being in this situation. 3. DETERMINANTS OF HEALTH A recent Ministry of Health Report (Social Inequalities in Health: New Zealand 1999, in press) looking into the relationship between social inequalities and health in New Zealand has found that inequalities in the basic distribution of material resources, income, education, employment and housing generate health inequalities. It has also been found that there is a there is a strong and consistent relationship between socioeconomic factors (and behavioural risk factors (such as smoking, and hazardous patterns of alcohol consumption and obesity). Likewise there is a strong and consistent relationship between socioeconomic factors and health outcomes such as poorer self-perceived health, hospitalisation, morbidity and premature mortality. Behavioural risk factors undoubtedly mediate some, but not all, of the association between socioeconomic factors and health outcomes. 8
It has been found that people who are better educated, have professional occupations, higher incomes and live in less socioeconomically deprived neighbourhoods are highly likely to enjoy better health and live longer than those who have no qualifications, are unemployed or in low-skilled jobs, earn less and live in socioeconomically deprived neighbourhoods. These health disparities however are not just evident across the extremes (those with the most compared with those with the least), but are apparent at every step between the extremes. There is thus a clear gradient in health disparities or inequalities. All sectors of society are affected, there are no neat cut-off points. As a result, health status is distributed as a gradient up the social hierarchy and applies to almost all causes of death from cancer, cardiovascular disease and Alzheimer s dementia to accidents and suicides. Individual health-related behaviours such as smoking only partially explain this strong graduated relationship, and, in any case, such behaviours are themselves socioeconomically patterned. Not only do these factors powerfully predict health and premature mortality at a given point in time, but this is also a lasting relationship these health inequalities persist over generations. Socioeconomic status is distributed inequitably across demographic groups in New Zealand. Most notably, the socioeconomic status of Maori and Pacific peoples is generally lower than that among New Zealand Europeans. Non-Maori are overrepresented in the least disadvantaged deciles of the NZDep96 deprivation index and Maori are over-represented in the most disadvantaged deciles. There is also an uneven distribution of socioeconomic resources by region of the country. Measures of average income and income inequality were calculated for 27 regions in this report: Wellington had the highest income, followed by Auckland City and then other areas in the Auckland, Wellington-Hutt and Waikato regions. Auckland City had the highest measure of income inequality. Wanganui was shown to have the lowest average individual income but also a relatively low measure of income inequality. Overall, potentially avoidable mortality increases with increasing deprivation and life expectancy at birth and from the teen-age years onwards declines markedly as small area deprivation increases If everyone in deprivation deciles 2 to 10 had the same mortality rates as those in the least deprived decile then, overall around 4800 deaths would have been averted in 1996, approximately 17% of all deaths in that year. It has also been found that regionally there is a statistically significant relationship with average equivalised income and mortality and hospitalisations. There is also evidence that there has been an increase in overcrowding in low income households as people double-up to lower housing costs per person. It has been found that people living in overcrowded houses were significantly more likely to perceive their health as poorer than those not living in crowded houses. 9
4. INCOME * 4.1 Wanganui District and All New Zealand compared It is apparent in comparing the distribution of Wanganui District household incomes at the time of the 1996 census with that of All New Zealand, that Wanganui households are distributed more to the lower end of the income range than are New Zealand households in general (see Figure 1). These are the households which reported the value of their household income in their census return (84 percent of New Zealand households responded; and 85 percent of Wanganui households). Incomes are Gross ; that is, before tax and including benefits and pensions. (It should be mentioned that a comparison of census incomes data with that from the Household Economic Survey suggests that census incomes are underreported. The Household Economic Survey has more information about different sources of income.) Figure 1 Comparison of distribution of 1996 census household incomes. (Wanganui District and New Zealand) 30% 25% Wanganui NZ 20% Percent 15% 10% 5% 0% < 15 15-20 20-25 25-30 30-40 40-50 50-70 70-100 100+ Census income bracket Mean and median household incomes calculated from these data are given in Table 1. These show that Wanganui average incomes are substantially lower than All New Zealand average incomes. To some extent these lower average incomes are offset by Income (as asked in the Household Economic Survey and census) is supposed to include cash income before-tax, from all sources, including benefits and the accommodation supplement. 10
lower average housing costs, and also by needing to spend less on average on some other consumption items, for example local transport. Table 1. Mean and Median 1996 Census Household Incomes: Wanganui and New Zealand. Wanganui New Zealand Number of households 16,773 1,276,332 (including non-respondents) Mean household income $34,705 $45,926 Median household income $27,021 $34,707 4.2 Relative Poverty and Equivalised Income Groups of people within Wanganui who will be most affected by an increase in rates, and thus a reduction in discretionary income, are those on lower incomes. Establishing who this includes is complex. There are a number of concepts employed by researchers to help capture the groups whom may be considered poor or on lower incomes. Two such notions used to determine this are relative poverty and equivalised income. Relative poverty is not easily defined. Often an arbitrary poverty line is drawn to serve as a measure of relative poverty. An example of this could be to define a household living on less than $15,000 per year (1996 prices) as poor. In which case, from the figure above, 24.5 percent of Wanganui households were poor at the time of the 1996 census, compared with 17.5 percent for New Zealand in total. In terms of individual persons in households reporting household income, 15.8 percent in Wanganui lived in households with income below $15,000 compared with 11.0 percent for New Zealand as a whole. There are however other factors which complicate matters further. Using income as described above does not take into account the individual circumstances of these households. An example of such circumstances could be superannuitants living alone in their mortgage-free home and whilst hardly having a luxurious standard of living, they would not normally be described as living in poverty. The standard approach to finding a more real measure of relative poverty is to calculate equivalised incomes, which take into account the number of persons living in the household and often, the composition of the household, such as the number of Census income data is reported in bands eg $10001 to 15000. Calculations for those above and below the poverty line have been made assuming each household in a given band had the mean income for that band. In reality, households will be spread over the whole band width and in some cases, those households in the upper part of a band which is allocated to the below poverty line group would in fact be above the poverty line. This means that the poverty line as defined does take in a proportion of households above the equivalised $15,000, pushing up the proportion shown to be in relative poverty. This would affect the NZ benchmark as well as the Wanganui numbers. 11
adults and children. An equivalence scale is applied to the household s actual income, converting it to an equivalent income which measures the standard of living of the household in terms of the income which would provide an equivalent standard of living for a two adult household. Thus the actual income of a two adult plus child household is scaled down to reflect the fact that the household s income has an extra person to support. The income of a person living alone is scaled up, because the income supports one person only. In general the scale will allow for economies of scale in household expenditure. That is, two people living together can live more cheaply than living separately. The most commonly used Equivalence Scale in New Zealand is the Revised Jensen Scale (RJS), which allows for the different spending needs of adults and children. However, a simpler scale the Luxembourg Income Study (0.5) Scale (LIS), which is increasingly being used overseas, simply adjusts income by dividing by the square root of the number of persons in the household. The LIS scale is used here as it requires less information than the RJS scale, and does not appear to give inferior results. Applying the LIS scale to the 1996 census household incomes data gives the results shown in Table 2 below. Table 2 Mean and Median Equivalised 1996 Census Household Incomes: Wanganui and New Zealand. Wanganui New Zealand Mean household income $31,980 $40,853 Median household income $25,681 $31,211 It is important to note that these equivalised incomes are not in actual dollars. They are an index of a household s standard of living. The poverty line at $15,000 is equivalised income. Relative poverty is generally defined by those working in the field in terms of equivalised incomes. A widely-used definition of relative poverty is that all those whose equivalised household income is less than 60 percent of the equivalised median household income are poor. See for instance Waldegrave et. al. (1999), and references therein. This would imply a poverty line for New Zealand of 60% of $31,211; or $18,727 (see Table 2). This is in terms of gross income, whereas poverty lines are more usually defined in terms of after-tax or disposable income; and would therefore be somewhat lower down the income distribution because taxes reduce median incomes more than low incomes. Also this poverty line does not allow for differences in housing costs. 12
An appropriate poverty line drawn from census equivalised gross household income would therefore be somewhat lower than $18,727; and even lower for a region with relatively low housing costs. As an ad hoc poverty line for this discussion we use a value of $15,000. That is, persons with an equivalised gross household income or less have a standard of living (to the extent that income determines standard of living) equal to or less than that of a couple with a combined pre-tax income of $15,000. Limitations of an ad hoc poverty line are that it does not allow for different incomes in different regions and also different expenditure needs for example lower housing costs in rural areas. To the extent that housing costs are lower in Wanganui than for All New Zealand and ad hoc poverty line could overstate the amount of poverty in Wanganui. On this ad hoc basis the proportion of households, and of individuals, living in poverty in Wanganui and in New Zealand are shown in Table 3. Table 3 Proportions of households and persons below the poverty line. Wanganui and New Zealand, 1996 census. Poverty line defined as $15,000 equivalised pre-tax household income. Wanganui New Zealand Households 16.7% 12.9% (number) (2394) Persons 16.9% 12.7% (number) (5931) Note: The numbers cover only those households reporting their household income; 85.3% for Wanganui. 4.3 Groups below the poverty line A rates increase, or a subsequent rents increase, would be expected to have its greatest impact on the households falling below a certain equivalised income as described above. According to the definition of relative poverty used here they account for about one in six of all households, and persons, living in Wanganui. A breakdown of some general characteristics of individuals living in these Wanganui households is shown in Table 5. The numbers in this table exclude those temporarily absent, and so will be a couple of hundred less than the total shown in the previous table. It should be noted also that the non-response rate to the census household income question is higher for Maori (about 25 percent for All New Zealand, as against 16 percent overall) so that Maori numbers are likely to be under-estimated. 13
Table 4 Characteristics of individuals living in Wanganui households below the 1996 census poverty line. Poverty line defined as $15,000 equivalised pre-tax household income. Number % of total % of this in poverty population group Ethnicity: European 3,735 64.9% 12.0% Maori 1,785 31.0% 23.7% Other 225 3.9% 19.4% Total 5,757 100.0% 14.5% Age: 0 15 2,481 43.1% 23.8% 16 24 687 11.9% 15.1% 25 44 1,428 24.8% 12.9% 45 64 750 13.0% 9.3% 65 + 414 7.2% 7.2% Household size: One persons 822 14.3% 20.2% Two 999 17.4% 9.5% Three 1,497 26.0% 20.0% Four 1,035 18.0% 12.1% Five 891 15.5% 16.0% Six + 513 8.9% 13.9% Note: Numbers do not add exactly because of census random rounding to multiples of 3. Source: 1996 Census On these numbers Europeans account for about two thirds of those in relative poverty in Wanganui, and Maori (broadly defined) about one third. However, the prevalence is higher for Maori with nearly a quarter being in relative poverty (and perhaps more if we had better coverage). Children account for 43 percent of those in relative poverty and nearly a quarter of all children live in a household below the poverty line used here. These will be predominantly children from sole parent households and other low-income family households. The largest proportion of those below the poverty line is found in three person households. The highest prevalence of relative poverty is for one person and three person households, each accounting for 20 percent of individuals in relative poverty. As discussed earlier, the poverty line used here to show the groups most likely to be affected is not an official and standardised measure. It is used, however, to give an indication of where poverty is most likely to be found. From the results above, it appears that poverty is most prevalent for low income families with dependent 14
children, including sole parent families. About a third of those beneath this poverty line are Maori. 5. IMPACT ON HOUSEHOLD BUDGET 5.1 Household Expenditure Patterns in New Zealand The data for this section are from Statistics New Zealand s Household Economic Survey (HES) for 1997/98. They are based on the survey responses of about 3,000 private households covering the whole of New Zealand. The accuracy of the estimates is limited by possible sampling variability. Household spending can be grouped into 7 broad categories, as shown in Table 5. Table 5 Household Expenditure by category all New Zealand private households. 1997/98. Category Av. Weekly Spending % of total per Household (adj. for capital outlays) Food $113.50 15.5% Housing 1 $177.20 24.2% Household operations $ 94.90 13.0% Apparel $ 26.50 3.6% Transportation $122.50 16.5% Other Goods $ 81.00 11.1% Other Services $116.50 15.9% Total 1 $732.20 100% Notes: 1. Adjusted by adding back capital outlays associated with housing (averaging $49.10 per week for all households). Source: Household Economic Survey (HES) 1997/98. Statistics NZ. Housing costs, which include rent, mortgage payments (interest and principal), rates, insurance, and maintenance are a substantial part of the average household s expenditure. Figure 2 shows how housing expenditure, as a proportion of total spending, varies over the range of household incomes. The categorisation of household income used here is the same as used in the 1996 census. (An alternative is the categorisation into 10 household income deciles, based on actual 1997/98 survey data. The census categories are used in preference, because of the use later in this report of some census material.) Spending on housing is around 25 percent of total spending over most of the income range. As will be seen shortly, the main variations are with household type, and by tenure of dwelling, rather than by household income. 15
Figure 2 Housing (adj*) expenditure as percent of total household expenditure by census income group. (All New Zealand households, 1997/98) 30% 25% 20% Percent 15% 10% 5% 0% 1 2 3 4 5 6 7 8 9 10 Census Income Groups * Adjusted by adding back capital outlays associated with housing (averaging $49.10 per week for all households). Figure 3 Expenditure breakdown by household census income groups: All New Zealand, 1997/98 100% 90% Other services Percent 80% 70% 60% 50% 40% Other goods Transport Apparel Hshld operations 30% 20% Housing 10% 0% 1 2 3 4 5 6 7 8 9 10 Census Income Groups Food 16
Note: Census income groups 1= <10001; 2=10001-15000; 3=15001-20000; 4=20001-25000; 5=25001-30000; 6=30001-40000; 7=40001=50000; 8=50001-70000; 9=70001-100000; 10=>100001. Figure 3 shows the composition of household expenditure by census income category. Spending on Other Goods and Other services tends to be a larger proportion of outlays for the higher income groups, whilst the proportion spent on Food and Household Operations tends to be higher for the lower income groups. Figure 4 shows Housing as a proportion of total spending, by household type. (A family type breakdown is also given in the HES tables.) Figure 4 Housing (adj.*) expenditure as percent of total household expenditure by household type. (All New Zealand, 1997/98) 40% 35% 30% 25% Percent 20% 15% 10% 5% 0% 1 2 3 4 5 6 7 8 9 10 Household types * Adjusted by adding back capital outlays associated with housing (averaging $49.10 per week for all households). The household types are: 1 One adult, not in full-time employment 2 One adult, in full-time employment 3 One adult, with one or more children 4 Two adults, neither in full-time employment 5 Two adults, one or both in full-time employment 6 Two adults and one child 7 Two adults and two children 8 Two adults and 3 or more children 9 Three or more adults without children 10 Three or more adults with children. 17
Children, in this categorisation are defined as aged under 15. The high proportionate outlays on housing for sole parent families are immediately apparent. Housing costs are also relatively high for couples with children. The lower ratio for category 4 is because of the concentration of retired couples in this category, many owning their home mortgage-free and with relatively low housing costs. Figure 5 gives housing cost proportions by tenure. Leaving aside the relatively small Rent-free category; those renting and those paying off a mortgage spend approximately equal proportions of their incomes about 30 percent - on housing. This includes rates, property insurance, and maintenance costs. There is of course a great deal of individual variation about these averages. Figure 5 Housing (Adj.*) expenditure as percent of total household expenditure by tenure type. (All New Zealand Households, 1997/98) 35% 30% 25% Percent 20% 15% 10% 5% 0% Rented Rent-free Owned with mortgage Owned without mortgage Tenure * Adjusted by adding back capital outlays associated with housing (averaging $49.10 per week for all households). Owned without Mortgage households on the other hand outlay only 13.4 percent of their total expenditure on housing costs. These households are in general the more mature households, either in the later stages of working life, or retired. Retired households have in general, of course, a substantially lower income. In summary, housing costs vary more with life-stage of the household, that is the age of the adults in the household and whether or not they have dependent children, than they do with household income. In particular, the kinds of household which are 18
most likely to be under pressure from housing costs are, from the above, the following households either renting or with a mortgage: Sole parent households (and presumably other beneficiary households). Couples with dependent children. 5.2 Household Expenditure Patterns in Wanganui District The HES sample of about 3,000 households is too small to give good direct estimates at District Council level. However, Statistics New Zealand have developed a regression model which links HES data to regional census data, and computes indirect regional estimates. Table 6 compares total expenditure on housing, and expenditure on all items, for Wanganui District with New Zealand. It is not appropriate to make similar comparisons of sub-components of housing expenditure, because the estimates are averages over all households, and do not take account of the different proportions renting, mortgaged, etc. Table 6 Comparison of Housing and Total Expenditure; Wanganui District and New Zealand. Estimated Annual Average Household Expenditure Housing All Items Housing as % of All WDC NZ WDC NZ WDC NZ 1995/96 $6,099 $6,547 $30,824 $34,364 19.8% 19.1% 1996/97 $5,601 $7,130 $29,831 $35,169 18.8% 20.3% 1997/98 $7,467 $6,706 $33,367 $35,754 22.4% 18.8% Source: Stats NZ estimates Housing costs would be expected to be lower on average for Wanganui than for All New Zealand. This is the case for the 1995/96 and 1996/97 estimates, but not for 1997/98. Almost certainly this is an estimation error generated by the regression model. The All items expenditure estimates are, however, consistently lower for Wanganui than for All New Zealand, as would be expected. 6. HOUSEHOLDS AT RISK OF BEING AFFECTED BY A RATES INCREASE From the prior discussion, it is apparent that the households most likely to be at risk of a change in discretionary income affecting health and wellbeing are low-income families with children, both sole parent and couple families, either renting or paying 19
off a mortgage. The majority of persons in such households are non-maori, but a relatively high proportion of all Maori live in such households. It is important to note that a quarter of those living under the poverty line are children. 6.1 The effect of relative poverty on household consumption. Little research has been carried out on the effects of relative poverty on household consumption. However, one useful account of how those on very low incomes cope is provided by a recent report by the Lower Hutt Family Centre (Waldegrave et. al., 1999). A total of 401 low income households were randomly selected in the 10 main centres, and face-to-face interviews conducted in June and July 1998. To be designated low income, households had to fall below the poverty line set by the Family Centre at 60 percent of median disposable (after-tax) household income, after taking housing costs into account. Any Accommodation Supplement received is treated as part of income. The Centre estimates that 18 to 20 percent of New Zealand households would be in this category. Specimen poverty lines for different household types were as follows: Household disposable income no greater than a. 1 Adult + 1 child $17,000 b. 2 Adults $18,000 c. 2 Adults + 2 children $26,000 (In terms of our approach to equivalising census household incomes given earlier, household types a and b would have equivalent income thresholds equal to $17,000 and $18,000; and household type c $13,000. These Family Centre thresholds are for after-tax income rather than before-tax income, and are applied to larger centres with higher housing costs.) A few of the key results from their detailed findings are as follows: 44% paid 40% or more of their income (after tax) on rent or a mortgage; and 25% paid half their income or more; 24% of all respondent said they could not afford to buy essential food items most times when they bought food; 56% of all households had members who did not visit a doctor when they needed to in the previous year because of a shortage of money. The study concluded that many low income New Zealand households are substantially deprived of essential household items and services. 20
6.2 The effect of a cut in income on household spending patterns. The best, virtually the only, source of information on what New Zealand households do in response to a change in their discretionary income is a recent analysis by Michelini (1999) published in New Zealand Economic Papers. Michelini used data from the Household Economic Survey for the nine-year period 1983/84 to 1991/92. Households were classified into five income groups and four household sizes. He analysed expenditure totals for six of the seven main expenditure categories: Food, Household Operations, Apparel, Transport, Other Goods, and Other Services. He omitted Housing because of the difficulties caused by inclusion of sale and purchase of property. The omission of Housing from his analyses is not a significant drawback for our purposes. Housing costs are a relatively fixed expenditure, which would not be expected to change much in response to a reduction in discretionary income, so it is reasonable to assume that almost all the impact of any income change will be felt in other categories of expenditure. Using this he calculated the elasticities of response for expenditure in each category to a change in total expenditure. Changes in total expenditure could be expected to match reasonably closely changes in discretionary income, more especially for those on lower incomes. The relevant values from Michelini are given in the column headed Expenditure elasticity in Table 7. To take an example, the value of 0.558 for the Food group means that a cut of 1% in a household s overall consumption budget would lead, on average, to a 0.558% cut in spending on food items. The value of 1.321 for Other services means that spending on that group would be reduced by 1.321% for an overall expenditure reduction of 1 percent. Table 7 Effect of reduction in overall household budget from an increase in housing costs Expenditure elasticity Share of nonhousing expenditure Effect of budget reduction of: $50 $100 $150 $800 Food 0.558 20.5% $5.71 $11.41 $17.12 $91.31 Household 0.747 17.1% $6.39 $12.78 $19.16 $102.20 operation Apparel 1.11 4.8% $2.65 $5.30 $7.95 $42.41 Transport 1.276 22.1% $14.08 $28.17 $42.25 $225.35 Other goods 1.095 14.6% $7.99 $15.98 $23.98 $127.87 Other services 1.321 21.0% $13.87 $27.73 $41.60 $221.87 Sum 100.0% $50.69 $101.38 $152.07 $811.01 21
Food and Household Operation (fuel and power, telephone, appliances, furniture, etc.) have elasticities less than one. That is, they are budget-inelastic. Spending on them changes less in response to a change in discretionary income than does spending on the other four groups including Apparel, Transport, Other goods, and Other services which have elasticities greater than one, that is are budget-elastic. These elasticities are for broad expenditure categories, combining elasticities for individual goods and services within each category. Elasticities for specific goods and services in each category could vary considerably from the values given in Table 7. Also the given values are calculated over all households, and any individual household might have a quite different response from the average. The second column in Table 7 shows the average 1997/98 share of each expenditure category in household expenditure, excluding expenditure on Housing. (Housing expenditure amounted to 18.8 percent of total spending on average during 1997/98.) From these first two columns the effect of a given reduction in the overall household budget on spending in each category can be estimated. The table gives the results of these calculations for amounts of $50, $100, and $150; and also for an amount of $800. The smaller amounts cover a range of possible rates/rents increases to meet the cost of upgrading Wanganui s wastewater scheme. The $800 represents a possible cost per private household of having to separate storm-water and sewage flow on their properties. To illustrate, a budget reduction of $100 per year is estimated to lead on average to a cut in spending on Food of $11.41 per year, on Household Operations of $12.78, and so on. The largest reductions are estimated to be for Transport and Other services. The total cuts over the six spending categories add to the $100 budget reduction. (In fact, they don t quite. This is because mathematically Michelini s model is calculated in terms of logarithms, and converting back to actual values leaves a small error in the sum of the calculated values.) Michelini s analyses allow us then to calculate the adjustments we expect households to make on average to their spending patterns as a result of a given change in their household budgets. We cannot say that a specified individual household will make precisely these adjustments, but we expect households on average to do so. 7. AT-RISK HOUSEHOLDS AND THEIR HEALTH Population health is known to be associated with areal deprivation, as measured for instance by the NZDep96 index of deprivation (Salmond, Crampton, and Sutton; 1998) already being used by Wanganui District Council in association with Good Health Wanganui. Household health outcomes are also known to be associated with the level of household income. This is illustrated for example by the two charts below from a recent report by the Wellington School of Medicine to the Ministry of Health, analysing data from the 1996/97 New Zealand Health Survey. The proportion of adult males and females reporting their state of health as either Excellent or Very Good (age-adjusted) increases significantly with income (gross household income, equivalised with the LIS scale). 22
Figure 6. Adult Females aged 15-64: State of Health Excellent or Very Good by Income Quintiles 100 80 Percent 60 40 20 0 1 2 3 4 5 Equivalised Household Income quintile NB where 1 is the most deprived and 5 the least deprived Figure 7 Adult Males aged 15-64: State of Health Excellent or Very Good by Income Quintiles 80 Percent 60 40 20 0 1 2 3 4 5 Equivalised Household Income quintile NB where 1 is the most deprived and 5 the least deprived This is not to claim that low income is in itself the basic cause of poor health. The mechanisms linking socio-economic status, income, and health outcomes are complex. But it is likely that low incomes do make some direct contribution to poor health for some households; for instance through inadequate nutrition and heating, substandard housing, and inability to afford health services when needed. The Family Centre report cited earlier shows that poor households could not always afford treatment when required. The untreated illnesses and conditions were generally minor, but in some instances more significant or potentially so (op. cit. Page 33). 23
7.1 Heating costs One aspect of household costs that have an impact on health, particularly of older people is expenditure on heating. Lack of disposable income and economising on fuel is a major determinant of the indoor temperature of older people s houses. For example in Britain there has been growing concern about fuel poverty and extra payments are made to older people to assist with heating costs (Howden-Chapman et al 1999). Although New Zealand has a temperate climate, there is greater seasonal mortality here than the more extreme northern European climates, particularly in those aged 65 and over (Isaacs & Don 1993). Cold stress, due to the inadequate construction of older New Zealand homes and inadequate heating seem to be the dominant factors responsible for increasing cardiovascular disease in those over 65 years. Although we do not have seasonal mortality data specifically for Wanganui, one risk of a rates increase could be that older people might economise on heating to the detriment of their health. 7.2 Food Security The Ministry of Health s 1997 National Nutrition Survey (Ministry of Health 1997) found that many aspects of food security were associated with deprivation. It was found that individuals living in the most relatively deprived quartile were more likely to report: their household could afford to eat properly only sometimes; food runs out in their household, often or sometimes; they eat less because of a lack of money; variety of foods limited because of lack of money; they were more likely to be in a position where they had to rely on others to provide food and/or money for food ; feeling stressed because of not having enough money for food; and feeling stressed because of not being able to provide the food they wanted for social occasions. These constraints on food are potentially problematic for the health of low-income families particularly as low-income families include a proportionally high number of children. 8. THE ACCOMMODATION SUPPLEMENT Low income households receive substantial public assistance with their housing costs, in the form of the Accommodation Supplement. Rents in excess of a specified threshold value are partly subsidised, up to a specified maximum, for those whose income is below a specified level which varies with family size and composition, and who also satisfy an asset test. The threshold and maximum values and income limits 24
vary geographically. There are three tiers of subsidy, the highest being the Area 1 rates for Auckland; then the Area 2 rates for Wellington, Hamilton, Rotorua, Napier, Hastings, Tauranga, Palmerston North, Nelson, and Christchurch; and finally Area 3 subsidy rates for the rest of the country, including Wanganui. The Area 3 subsidy limits applying to Wanganui for various household types are as follows (from the current WINZ brochure): Table 8. The Accommodation Supplement. Income limits and accommodation cost thresholds. Income Bottom threshold for weekly ceiling accommodation costs (before tax) Rent Owned Single (18+) with no children $21,799.44 $46 $56 Single with 1 child $27,715.48 $73 $87 Single with 2+ children $33,061.60 $77 $93 A couple with no children $32,033.04 $77 $93 A couple with children $36,193.04 $89 $107 The Accommodation Supplement pays, for those meeting the income test (and asset test, described below), 70 cents of each dollar by which accommodation costs exceed the thresholds tabled above. This is up to a maximum of $75 per week, or until the Accommodation Supplement takes income above the Income ceiling, whichever comes first. Assets are also tested (income-generating assets, excluding the property in which the family is resident). If assets exceed $8,100 for a single person with no children, or $16,200 for a couple or sole parent, no supplement is payable. Assets under these limits but in excess of specified thresholds ($2,700 for a single person with no children; $5,400 for others) are added to income at a conversion rate of $1 a week income per $100 of assets. The allowable accommodation costs for low-income home-owning households include mortgage costs (interest and principal repayment), insurance (including mortgage protection insurance, but excluding contents insurance), rates, and major repairs. Boarders can also qualify for the Accommodation Supplement. The rent component of their boarding costs is assessed at 62 percent. Total Accommodation Supplement payments, nation-wide, are substantial of the order of $700 million per year assisting about 300,000 households, or over $2,000 per average subsidised household. Most recipients of the supplement are renters and boarders; 178,154 and 73,876 respectively in 1997, compared with 45,071 mortgagors. The latter were 15.2 percent of the total (DSW 1997). 25
The implication is that low-income Wanganui households facing an increase in housing costs will not necessarily have to meet the total cost from their own resources. If they are not already up to the subsidy ceiling they will have to meet only 30 percent of the increase. If they are already at the ceiling, they will have to meet the full cost. Note that not all qualifying households actually take up the supplement and we are not aware of any estimates of uptake that could indicate the extent to which those entitled to the supplement do not know they are entitled or do not how to get it. These numbers could be significant. We neither have information on the number of Wanganui households currently receiving the Accommodation Supplement, nor of these how many have reached their subsidy ceiling. Provided households are below the income limit they can qualify for the Accommodation Supplement for accommodation costs ranging from the bottom thresholds given in the above table up to a ceiling of: threshold plus $75 divided by 0.70. For the parameters given in the table above this would typically be up to a ceiling of $180 + per week. Average rentals reported in the 1996 census were well below these upper limits, though rents will have increased a few percent since 1996, and of course some variation is to be expected about the averages. Figure 8, however, shows that there is not a lot of variation, at least in average rents, by census area unit within the urban part of Wanganui District. Figure 8 Wanganui City Area Units: Average rents against deprivation. (excluding rural area units) $180 $160 Avearge weekly rental - 1996 census $140 $120 $100 $80 $60 $40 $20 $0 1 2 3 4 5 6 7 8 9 10 1996 Index of Deprivation - from 1 (low) to 10 (high deprivation) 26
Average weekly rents by Area Unit, in 1996, were almost all in the $100 to $140 per week range. Of the two notable outliers, the low average was for Putiki, and the high for Spriggens Park (however, these outlier values could be affected by the statistics New Zealand practice of randomly rounding numbers of dwellings to a multiple of 3). Rented properties tend to be more concentrated in those parts of urban Wanganui with high levels of socioeconomic deprivation (those with index scores of 9 or greater included Balgownie, Castlecliff, Gonville West, Mosston, and Laird Park). Approximately 53 percent of the 3,807 rental private properties in the 1996 census were in urban Area Units with a Deprivation score of 8 or higher (but included only about 40 percent of the total population). 27
9. CONCLUSIONS A reduction in discretionary income will increase hardship for a minority of Wangaui households. This may be mitigated somewhat by entitlement to the Accommodation Supplement which can meet 70 percent of the extra cost for low income households. There is however, no evidence as to the uptake of the supplement. Those who do not know they are entitled or do not know how to apply are likely to be the most at risk. Categories of households most likely to be at risk are low-income families with children, both sole parent and couple families, either renting or paying off a mortgage. The majority of persons in such households are non-maori, but a relatively high proportion of all Maori live in such households. A quarter of children within Wanganui live below a poverty line set at $15,000. Estimates are given in this report of how any reduction in household discretionary income would be allocated over spending categories. For the average household spending is cut back in all categories, but less so for Food, and Household Operations. Individual households respond each in their own way to a cut in their income. It is difficult to predict which, if any, will adjust their consumption in ways which might threaten their physical or mental health. For most households a rates increase of $100 per year would not be a significant threat to well-being. This however, cannot be said for all households. Wanganui households will of course share the environmental benefits of the upgraded wastewater scheme. The physical health of such families could be at risk by insufficient spending on nutritious food, on heating, and on medical care. A recent Family Centre report on low income families shows that many have difficulty in keeping up essential outlays on all these items. Psychological health is more difficult to relate to consumption choices, but poor families do have difficulty in participating in many social activities. New Zealand research has found a strong and consistent relationship between socioeconomic factors (such as deprivation, education, employment, income, occupational class and housing conditions) and health outcomes such as poorer selfperceived health, hospitalisation, morbidity and premature mortality. It has also been found that people who live in less socioeconomically deprived neighbourhoods are highly likely to enjoy better health and live longer than those who live in more socioeconomically deprived neighbourhoods. Compared to All New Zealand, Wanganui has a higher proportion of people living within socioeconomically deprived neighbourhoods. 28
REFERENCES BERL (Business and Economic Research Ltd) April 2000. Historical Economic Profile. Report prepared for Wanganui District Council by Sarah Russell and Ganesh Nana. Department of Social Welfare. Statistics Report. Wellington. 1997. Howden-Chapman, P., Signal, L. & Crane, J. Housing and Health in Older People: Ageing in Place. Social Policy Journal of New Zealand, 1999, 11, 14-30. Howden-Chapman P (ed). 1999 Social Inequalities in Health: in New Zealand 1999. A report to the Ministry of Health (to be published). Wellington School of Medicine. Isaacs, N. & Donn, M. Housing and Health seasonality in New Zealand mortality. Australian journal of Public Health, 1993, 17, 68-70. Michelini C (1999). New Zealand Household Consumption Patterns 1983-1992: An Application of the Almost-Ideal-Demand System.. 33(2), December 1999. Pp. 15-26. Ministry of Health, 1999. Taking the Pulse. The 1996/97 New Zealand Health Survey. Statistics New Zealand. Household Economic Survey. Annual until 1997/98. The 1997/98 summary tables are on website www.stats.govt.nz. Waldegrave C, King P, and Stuart S (1999). The Monetary Constraints and Consumer Behaviour in New Zealand Low Income Households. The Family Centre Social Policy Research Unit. PO Box 31-050, Lower Hutt. 29
APPENDIX 1 TERMS OF REFERENCE 1. Introduction The Wanganui District Council is commissioning the research and analysis of a series of questions in relation to the impact on individuals and the community of increasing rates, to pay for a much needed upgrade of Wanganui s wastewater scheme. The Council has invited the Department of Public Health at the Wellington School of Medicine to submit a proposal to undertake three of these questions. 1.1 Background The current wastewater scheme is a significant financial burden on Wanganui ratepayers. Major environmental gains have already been achieved since the project s inception. Sewerage no longer enters the Wanganui River, except during some flood events. Council is working on modifying the wastewater scheme so that the additional environmental benefits for the community should become affordable by ratepayers for the community. The expenditure in Council s 10 year Plan reflects a scenario that involves the treatment of effluent in aerated lagoons, followed by clarification, then ultraviolet disinfection and disposal through the sea outfall. The project will also involve the separation of stormwater drains from sewers. The wastewater scheme review is now well under way, with improved and more affordable scheme details having been developed. The application for the review is scheduled to be submitted to horizons.mw (Regional Council) in October 2000. The current resource consent issued in 1992 states: If the economic circumstances (excluding asset sales) of the Holder improves due to any overall improvement in the economic circumstances of the Wanganui Community to an extent that the Holder is in a position to increase the financial contribution to the project, the Consent Authorities may review the conditions attached to these permits pursuant to section 128 (a) (iii) of the Resource Management Act for the purpose of having the scheme or parts of the scheme presented sooner. The Wanganui District Council has already commissioned BERL to describe the historical economic profile of the Wanganui District over the ten year period (1990-1999). Their main findings were that Wanganui was consistent with the country as a whole, in that it showed improvements in almost all economic indicators between 1992 and 1999. The analysis however, shows that the improvement for Wanganui was considerably less than for the rest of the country (BERL 2000). It was also shown by the Community Affordability Measure (CAM), which provides an indication of the relative affordability of living in a particular area, that between 30
1991 and 1996 affordability in Wanganui improved relative to the rest of the country. This can be attributed to the small rise in house prices during this period. If the CAM is measured between 1986 and 1996 however, relative affordability in Wanganui worsened substantially. This can be attributed to much smaller increases in income in Wanganui during the 1986 to 1991 period. 2. Description of proposed services The three questions the Wanganui District Council is seeking to be answered by the Wellington School of Medicine are the following: Question 5: Question 6: Question 7: Is a decrease in household discretionary income likely to result in a decline of expenditure in those goods and services required to maintain the physical and mental health? What is the impact on household s financial capacity to maintain the physical and mental health status if rates are increased in real terms by say $50 or $100 or $150? Under the existing rating system what is the Affordability Threshold for the Wanganui community before risk of significant harm eventuates? Target households to be investigated: Our method would be to look predominantly at: Pensioner households, sole person or couple; Beneficiary households; and Other low-income households. Other low-income households will be more difficult to define as income in this context means income relative to needs. Thus a large household with many children might be thought of as low-income even though actual income is relatively high. Renting households and owner-occupied households will also be included in our analyses as increases in rates tend to be passed on through rents. However, where data are available, we will endeavour to distinguish between renters from owners. Data Sources: Main data sources: Household Expenditure Survey (HES), giving all New Zealand expenditure averages, the latest year being 1997/98. 1996 census, giving household income information, by geographical area. (Wanganui district council, or the Manawatu-Wanganui Regional Council may already have detailed census tabulations? It would be helpful if tables 31
were available of household income by household type and income source, and other household characteristics.) An article in a recent issue of NZ Economic Papers (December 1999), by Michelini, estimating for broad household expenditure groups the spending response to changes in available income. This work is based on HES, and so the averages are for all New Zealand, and also for all households combined. However, it should give some idea of the likely responses by different household types to budget changes. Other potential sources include: A report by the Lower Hutt Family Centre on responses to the 1991 benefit cuts. Recent research using the 1996/97 National Nutrition Survey, assessing family food security. Recent work at the Wellington School of Medicine on the relationship between household income and health outcomes. Overseas research on the effect of reduced household heating on mortality. Proposed Research Approach: We cannot provide precise answers to the questions posed by the Council. We can however, give a general indication of likely effects based on existing data and research. A reduction in discretionary income will result in less spending, but households will differ considerably on where they make the cuts, and whether the cuts they make are sufficient to place their physical and mental health at risk. The larger the increase in rates, the more likely it is that some households will be at risk, but it would be extremely difficult to say how many. Information however can be provided on how average households might be expected to behave. We propose to: (h) Tabulate average all New Zealand household spending patterns for selected household types from HES. (i) Make assumptions about extent to which these might need adjusting to better fit Wanganui households for example lower housing costs. (j) Using census material, make estimates of the number of low-income households of various types in Wanganui district. (k) Calculate, using the Michelini estimates, the likely consequences of reductions of the given magnitudes ($50, $100, $150) in discretionary income. (l) Cross-check with Family Centre results to the extent possible. (m) Tabulate also some of the material referred to above on the relationship between household income and health. 32
(n) Discuss and report on the implications of the material in (a) to (f), and also that on nutrition and heating. 33
APPENDIX 2