Assessing the progress on poverty reduction.

Abstract

This paper presents the summary results of the New Zealand Poverty Measurement Project's analysis of the incidence and severity of poverty during the 1990s, and assesses the impact of five social and economic policies introduced by the Labour-led coalition governments since 1999: New Zealand superannuation, income-related rents on state houses, active labour market policies that promote an employment-rich economy and reduce unemployment, the Primary Health Care Strategy and the planned income support policy to reduce child poverty. Superannuation is assessed as both adequate and sustainable, rents for state houses are found to be affordable, and GDP growth and employment have increased incrementally as unemployment and benefit numbers have decreased. The Primary Health Care Strategy is seen as an innovative initiative that will increase affordable access to general practitioners, but it and the proposed child support initiatives are too new to be adequately assessed. Of the challenges that remain, policy priorities should centre on housing alternatives, including home ownership for low-income households not in state houses; income support for poor households, particularly those with children; and multi-sector development of education and training aimed at lifting economic and social capacity.

INTRODUCTION

In 1991 New Zealand's Finance Minister, Ruth Richardson, announced a Budget that she signalled would be "the Mother of All Budgets" (Louisson 1991), featuring benefit cuts, market rents on state houses and the introduction of a range of new user charges. The Budget led to difficulties for low-income households, which were already under some pressure from a period of rising unemployment and economic restructuring. In 1999, the Labour-Alliance coalition came into office promising new social policies that would begin a process of substantial poverty reduction in certain key policy areas in Aotearoa New Zealand.

This paper presents the summary results of the New Zealand Poverty Measurement Project's (NZPMP) analysis of the incidence and severity of poverty during the 1990s and assesses the available evidence of the impacts on that poverty of the social and economic policies introduced by the Labour-led coalition governments since 1999. (2)

POVERTY MEASUREMENT DURING THE 1990s

Most modern definitions of poverty in OECD countries are relative in the sense that they relate to the living standards of that society. The definition of poverty adopted by the NZPMP is also a relative one: poverty is a lack of access to sufficient economic and social resources that would allow a minimum adequate standard of living and participation in that society.

Poverty is and always will be a contested concept. Even when there is broad agreement on a high-level definition there remains considerable debate over how best to measure poverty. One of the main reasons for the lack of consensus is the need for judgements to be made as to what constitutes a minimum adequate income or a minimum adequate standard of living. (3) As neither the income nor the living standards data can tell the researcher or analyst where to "draw the line", some external way to assess adequacy is needed. The NZPMP's measure of poverty is income based, but in contrast to approaches that simply set a poverty line at an arbitrary fraction of the mean or median household income, the NZPMP has sought to address the issue of assessing adequacy in an explicit and transparent way. The approach involves the use of focus groups to draw on the knowledge and practical experience of low-income householders (i.e. on their judgement) to estimate minimum adequate household expenditure in a full range of household expenditure categories.

In 1992 the NZPMP began a comprehensive research programme into poverty measurement. (4) From 1993 to the present, NZPMP has undertaken ongoing focus group sampling of low-income householders, in differing regions, cultural groups, family structures and employment categories, in urban areas, middle-sized cities and small towns throughout New Zealand, seeking information about minimum adequate budgets. The total weekly household estimate is intended to be minimal, but sufficient to live on independently--without having to resort to a food bank or Special Benefit. Furthermore, it is grounded in the everyday experiences of low-income households. These estimates have proved to be remarkably similar in the same regions and years. They have also proved to be sensitive to economic changes in regions (Waldegrave et al. 1996). From these resource-based estimates, an income poverty threshold was set--a realistic poverty line for use in social and economic policy.

The unit record data in the Household Economic Survey (HES) have been used to extract national quantitative data on the numbers in poverty, the types of households involved and the depth of poverty (Stephens et al. 1995, 2000, Stephens and Waldegrave 2001, Waldegrave et al. 1996, 1997, Waldegrave and Stephens 2000, Waldegrave 2000). Surveys and numerous qualitative studies have sought information on the consumer behaviour, methods of budgeting, survival strategies and unaffordable expenditures of low-income households.

The NZPMP produced a number of measures of poverty in New Zealand, but set the poverty threshold as 60% of median, equivalent, disposable, household income. This level emerged from the focus group sampling of low-income households. (5) The NZPMP produced data both before and after housing costs had been paid, and the preferred focal measure was 60% of median, equivalent, disposable, household income after adjusting for housing costs. This "relative" measure of poverty emerged from the "absolute" assessments of minimum adequate budgets by the low-income householders in the focus groups. (6)

Using this threshold, the quantitative data indicated that between 1993 and 1998 around 19% of households (7) were below the poverty line. The data enabled us to identify key groups who were adversely affected and the dollar impact of housing costs. Around a third of the children and over 70% of the single-parent households fell below the line. Maori were more than twice as likely, and Pacific people more than three times as likely, as Pakeha (Europeans) to be in poverty. Nevertheless Pakeha households made up more than 60% of those who were poor. People not employed accounted for around 70% of those in poverty. Housing costs accounted for at least 60% of the income shortfall, experienced by households below the poverty threshold. On the positive side, superannuitants were above the poverty line in most years, suggesting superannuation was the only government transfer that kept people out of poverty.

As Table 1 shows, in 1993 and 1998 the numbers of households and individuals in poverty changed very little. The poverty-reducing impact of the governments' combination of tax and benefit policies became less effective, falling from 58% to 43%. They appeared to be more effective, however, at reducing the severity of poverty, moving from 51% to 81% effective, but this was largely due to a number of superannuitants dropping only just below the threshold that year. The dollars below the poverty line, which is the addition of the shortfall of all those below the line, increased 22% to over one billion dollars. The final column, referring to the poverty gap, shows there was a slight fall in the average severity of poverty. Again, this was influenced by the superannuitants.

The impact of housing costs on poverty can be calculated by comparing measures of the incidence and depth of poverty at the same percentile before and after paying housing costs. This method has been adopted by the British Department of Statistics in their "Below Average Income Series" (Department of Social Security 1993). It enables another snapshot of the impact of housing costs on the budgets of low-income families.

As Table 2 shows, the incidence of poverty in 1998, using the 60% of median disposable income, jumps from 15.4% before adjusting for housing costs to 19.3% after adjusting the poverty threshold for average housing costs. The poverty gap (i.e. the numbers of dollars by which poor people fell short of the poverty threshold) increases from $400.7 million to $1,010.1 million. The substantial impact of housing costs pushed more low-income households into poverty, and those already in poverty deeper into it. This latter calculation of the leap from $400 million to $1,010 million suggests that housing costs made up at least 60% of the income shortfall for poor families.

It is very interesting to note that after housing costs had been paid, most of the increase in poverty was due to public and private renters paying open-market rents. Thus, despite the operation of the targeted Accommodation Supplement, over 70% of state tenants and a third of private renters were poor after housing costs. The numbers in both sectors almost doubled. The structure of poverty changed, with increasing numbers of Housing NZ, private rental and mortgagor households, while reducing freehold owners many of whom were elderly. The poverty gap more than quadrupled for state house tenants and trebled for private renters.

To verify that these income-based poverty measures were likely to lead to poor living standards, a range of other quantitative and qualitative data was collected through a series of studies involving survey methods, focus groups and in-depth interviews over the decade. They identified serious problems of affordability and a lack of access to key services. As the quantitative data showed, housing costs often squeezed out other necessary costs, like adequate food, clothing and healthcare. The focus group work generated information concerning people's coping strategies, budgetary choices, and affordable and unaffordable items and services. The results of a national random survey of households below the poverty threshold carried out by the NZPMP typifies the data that were collected, identifying specific problems and shortfalls in the areas of housing, food and health (Waldegrave et al. 1999). (8) The households surveyed were all among the bottom 20% of New Zealand household incomes, and most had been unable to obtain essential food items, doctor's visits and prescribed medications due to shortages of money over recent periods.

Throughout the period following the 1991 Budget reforms, the NZPMP researchers identified the need for poverty-alleviating policies that focused on income adequacy, affordable housing, access to healthcare and sustainable employment. (9) By 1998, however, as Tables 1 and 2 show, poverty levels were as high, or slightly higher, than they were in 1993. After that, sickness benefits were reduced to the lower level of an unemployment benefit, and superannuation payments for an elderly couple began a process of being lowered from a floor of 65% of the average wage to 60% in 1999. At the time, the NZPMP criticised these policies, and pointed out that the changes to superannuation would eventually place below the poverty line one-half to two-thirds of the elderly who were currently above it.

ANTI-POVERTY POLICIES IN THE NEW MILLENNIUM

As noted earlier, the 1999 Labour--Alliance coalition government came to power on a platform of social policies that would begin a process of substantial poverty reduction in certain key policy areas. The Labour--Progressive coalition government likewise promised similar policies during the 2002 election. The rest of this paper will analyse and assess the five key new antipoverty policies that marked these two administrations. Three are already fully enacted, one is partially enacted and the fifth will be enacted next year.

New Zealand Superannuation

On 1 April 1999 the floor of superannuation and its relationship to the net average wage for an elderly couple began a process of reducing from 65% to 60% as one of the final social policies of the retiring National Government. (10) As noted above, the data from the NZPMP indicated that the policy would eventually push a majority of the elderly into poverty. The new Government lifted superannuation above the previous 65% level on 1 April 2000. They also established the New Zealand Superannuation Fund to ensure the level was sustainable over the next decades, as the numbers of elderly are expected to more than double by 2050.

The New Zealand Superannuation Act was eventually passed in 2001, stating, "The standard weekly amount of New Zealand superannuation (after the deduction of standard tax) payable to a married couple, both of whom are qualified to receive New Zealand superannuation, is not less than 65% or more than 72.5% of the average ordinary time weekly earnings" (s16a). A single person who is granted a living-alone payment receives 65% of the couple's rate; a single person not receiving a living-alone payment is entitled to 60% of the couple's rate.

In order to make these payments sustainable over the next decades, the Government is partially pre-funding superannuation by lifting the contribution from taxation from 4% to 6% of GDP. This is designed to smooth the costs over time (McCulloch and Frances 2001).

Income-Related Rents on State Houses

The 1991 Budget announced that tenants in state houses would cease to pay an income-related rent, which at that time was set at 25% of household income for a beneficiary or a beneficiary-level income. The Government of the day eventually passed the Housing Restructuring Act 1992, which among other provisions introduced market rents for tenants in state houses. An Accommodation Supplement was made available to tenants through the Department of Work and Income. As the earlier research showed, however, tenants' outgoings for renting state houses increased substantially, despite the Accommodation Supplement. (11)

The Labour--Alliance coalition government reintroduced income-related rents for state houses on 1 December 2000 at the same base proportion as before--25% of household income.

Promoting an Employment-Rich Economy and Reducing Unemployment

The new Government's strategy was to reduce unemployment through sustained economic growth, a new regional economic focus, the development of a more highly skilled workforce, an "employment-rich" environment, labour relations reform to better protect job security and wage bargaining, a rise in the minimum wage, and redistribution to disadvantaged groups. The overall aim was to lift the performance of the New Zealand economy and, through an active labour market strategy, enable those currently unemployed and those at risk of being unemployed to participate in the market. Employment was seen as a major route out of poverty.

The Primary Health Care Strategy

The Primary Health Care Strategy identifies the need to reduce inequalities and improve the health of all citizens, recognising that access to first-contact primary care services, particularly general practitioners, is often prohibited by financial barriers.

Over the next eight to 10 years Public Health Organisations (PHOs) are being phased in as the Community Services Card is phased out. During the transition period two funding formulae are being employed. An "Access" formula is currently used where medical practitioners form themselves into PHOs in low-income high-health-need areas. This formula is population based and ensures patients are charged either low fees or no fees at all. The "Interim" formula continues the use of the Community Services Card. Eventually, the Interim formula will be phased out as the Access formula becomes a universal means of primary health care delivery. Weightings that recognise the particular needs of deprived groups, cultural disadvantage, age and gender attract extra funding. These are designed to promote greater access for low-income and high-need populations. PHOs are also expected to deliver or arrange the delivery of a range of health promotion activities. The Government intends to set PHO budgets for pharmaceuticals and laboratory tests on a needs-based formula.

Income Support to Reduce Child Poverty

Components of the benefit system that provide income transfers for children have a systemic problem and have been falling behind inflation. Unlike the base benefit, Family Support (that part of the benefit system that provides payments for children) is not indexed to inflation. The rates have not been adjusted since 1998 and as a result the real value has declined by 5.5% (MSD 2002: part 2: chapter 1), reducing the overall value of benefit payments to households with children. The same is true for the Child Tax Credit, the Family Tax Credit and the Parental Tax Credit. (12) (This group of tax credits are often confusingly referred to as the Family Tax Credits.) As the NZPMP research showed, around a third of New Zealand children were below the poverty threshold.

The Governor General, in her Speech from the Throne outlining the new Government's policies during the opening of the current Parliament, indicated that there will be a mixture of improved income support and some form of indexation against inflation during this term of government. In the Address-in-Reply debate, the Prime Minister underlined the same commitment regarding income support for children in low-income families during this term of office (Clark 2002).

ASSESSING THE POLICIES

These five anti-poverty policies form the basis of the two Labour-led coalition governments' redistributive strategies since the 1999 election. They mark social policy for them, as the 1991 Budget marked it for the National Government. They set out their overall objectives of social cohesion as they relate to an integrated programme of social assistance. It is still early days to fully assess the effectiveness of these policies, but there are data that can provide a preliminary understanding and partial assessment at this stage. (13)

New Zealand Superannuation

The Shipley Government's new formula for setting the level of superannuation on 1 April 1999 pushed it substantially below the previous threshold of 65% of the average ordinary-time weekly earnings. The adjustment was linked to prices rather than wages, resulting in New Zealand superannuation falling to 62.66% of the average wage and continuing to fall to 60%. The Clark Government restored the superannuation payment levels on 1 April 2000, more than a year before passing the Act in 2001. (14) In line with their new formula of not less than 65% and not more than 72.5% of the average ordinary time weekly earnings, they set it at 67.73%. Couples received a $21.42 weekly increase and singles a $12.86 or $12.36 increase, depending on whether or not they were entitled to a living-alone allowance (MSD 2000).

The most recent data sets of the NZPMP provide a statistical basis to assess the effectiveness of the new superannuation payments as an anti-poverty strategy. In Table 3 both the numbers and levels of poverty for children, adults and seniors are set out, with calculations taken before and after housing costs. During 1997, superannuation was paid out at no less than the 65% level. During the 2000 year three months were paid at the 62.66% level and nine months at the 67.73% level. It is unfortunate that Statistics New Zealand ceased running their Household Economic Surveys (HES) on an annual basis after the 1997/98 year. Had the 1999 year been available, this table could have illustrated the impact of the reduction in superannuation payments that year.

The two columns entitled Adults 65+ highlight the impact of superannuation. Before taxation was deducted and superannuation was paid, 76.9% (1997) and 71.2% (2000) of elderly households would have been in poverty. However, after the tax and super payments, those in poverty reduced substantially to 30% (15) and 17.9%, respectively. The Government's effectiveness at reducing poverty through its tax and superannuation mix was therefore 61% and 74.9%, respectively. These figures need to be seen in relation to the second efficiency row, which shows that those below the line were only just below it. In terms of income, those below the poverty threshold on average were only 3.6% (96.4% of threshold amount) below it in 1997 and only 1% (99%) in 2000. If the Government was to make up the shortfall for those below the poverty line, it would have cost them $3.226 billion and $3.44 billion respectively in terms of market income. However, after tax deductions and superannuation payments, the make-up amount reduces to $117 million and $34 million respectively.

The same analysis can be followed after housing costs have been paid. In the case of the elderly, this reduces the numbers in poverty considerably to 12.3% in 1997 and 10.8% in 2000. This is because a large proportion of elderly people have lower housing costs after paying off their mortgages. A look at the disposable and efficiency rows across the whole table shows just how effective superannuation has been in lifting elderly households out of poverty when compared with the results for children and adults. It is also interesting to note how particularly effective the new superannuation was in the year 2000, despite it not being introduced until the fourth month in the year.

Around 70% of all elderly substantially depend on superannuation to survive adequately. Currently 18% (74,900 persons) of the elderly live on superannuation alone, while 52% (218,600 persons) live on superannuation and less than $5,001 per annum from other sources (SNZ 2001b).

It is important that superannuation is both adequate and sustainable over time. The data presented here suggest it is adequate. We should note though, the high proportion of mortgage-free housing among the elderly substantially reduces costs for this group as Table 3 shows, and if this were to change substantially a different result could be expected. Nevertheless, if the HES data had been able to capture a full year of payments, they would almost certainly have shown that all recipients were above the threshold. The Living Standards Research initiated by the Super 2000 Taskforce focused primarily on the adequacy from a broad perspective of material wellbeing. They concluded that:

   Although the population of older people is characterised by a
   relatively modest current income, the great majority of older people
   are not subject to appreciable levels of hardship or material
   deprivation. (Fergusson et al. 2001:141)

They also referred to the high level of mortgage-free housing among the elderly as making a contribution to this result.

Sustainability over time is a particular challenge regarding income for the elderly, because their numbers as a percentage of the population are predicted to more than double over the next five decades. Currently there are 471,860 persons over 65 years in New Zealand and they make up 11.9% of the population (31 December 2002, SNZ 2003c). Demographic projections indicate their numbers are expected to swell to 1,182,000 by 2051 and make up 25.5% of the population (SNZ 2001c).

As noted earlier, the New Zealand Superannuation Act 2001 requires the setting up of the New Zealand Superannuation Fund to ensure the current levels are sustainable over the next decades. This involves partially pre-funding superannuation by lifting the contribution from taxation from 4% to 6% of GDP. Initially the fund will finance payments to both current recipients and future superannuitants. By around 2025 the invested fund will be contributing to the growing numbers of superannuitants. This has been designed to smooth the costs over time (McCulloch and Frances 2001).

Income-Related Rents on State Houses

The restoration of income-related rents to a maximum of 25% of household income for those on a benefit or benefit level income was one of the major social policy initiatives of the 1999-2002 Labour--Alliance coalition government. As at 31 March 2003 there were 62,624 state house rental properties, of which 89% (54,554) qualified for an income-related rent. The average rent paid for these properties at that time was $85.28, less than half the market value (HNZC 2003b). The budgeted extra cost to the Government for the income-related rents initiative from 1 December 2000 when they were introduced to 30 June 2003 was $257 million (Treasury 2000).

For those on low incomes in state houses, this policy change has substantially reduced the level of poverty they would have otherwise experienced. When the scheme was introduced in the Auckland region, households were around $50 a week better off. In the Wellington region the), were around $35 better off and in Christchurch around $30. Even in the less urban areas, the savings were significant. For example, the average household weekly savings on the East Coast were $25, Otaki-Levin $21, Taranaki-King Country $22 and the South Island West Coast $17 (Gosche 2000).

Of course, a household had to be resident in a state house to qualify for the income-related rent, and the stock of state houses was reduced by around 10,000 by the National governments and their coalition partners during the 1990s. Nevertheless, the 54,554 households that qualified for an income-related rent consisted of 159,726 New Zealanders (HNZC 2003b). This was not an inconsequential number of some of the country's most needy people. (16) As an anti-poverty intervention, the income-related rents policy provided state house tenants with improved housing stability and greater discretionary spending. NZPMP showed in Table 2 that 37% of those in HNZ rentals in 1998 were below the poverty threshold.

Tenure stability has increased, probably as a downstream effect of the policy change. The average percentage national property turnover for HNZC properties dropped eight percentage points from 23% 31 December 2000 to 14.4% 31 March 2003 (HNZC 2003b). This suggests tenants are experiencing greater housing security and predictability.

Alongside the income-related rent regime, a substantial house acquisition and upgrading programme has been introduced to begin the restoration of the state house stock. During the current year HNZC is targeting a net increase of 886 houses nationally, 680 of which will be located in Auckland. Over the 2002 to 2006 period the Government has committed to acquiring over 3,000 more houses; 2,500 will be added directly to the state house stock, almost entirely in Auckland, and 500 will be acquired in Northland, the East Coast and Eastern Bay of Plenty to address sub-standard housing in those regions. Of these, 290 will be new state houses and the rest will come from concessionary loans to iwi and community groups, enabling them to develop their own low-cost rentals and rent-to-buy schemes. Fifty more community houses will also be added to the stock (HNZC 2003a).

The "Healthy Housing" programme was designed as a specific anti-poverty measure to reduce the spread of infectious diseases like meningococcal disease. In New Zealand, 30% of the admissions for infectious diseases are children. Overcrowding in poor households and the poor quality of some houses are considered to be major causes. The Healthy Housing programme involves extending existing houses for large families, adding bathrooms, bedrooms and other living space, and improving ventilation and insulation. In the 18-month pilot period to June 2002, work had been completed on 619 properties, including 119 that had bathrooms added. In the 2002/03 financial year the annual target is for 76 extensions and nine design improvements (HNZC 2002a). This is a "non-silo", innovative way of addressing serious health, housing and poverty problems, which indicates a significant new approach to policy implementation in this area.

The income-related rents policy, however, did not deal with all those in housing need. The waiting list for state houses is currently 11,627, of which 4,587 have been assessed as being in the top two categories of severe or significant persistent housing need (HNZC 2003a). There were also 280,706 households receiving the Accommodation Supplement in December 2002 (Work and Income 2003b) because their housing outgoings, either renting or mortgaging, are considered to be disproportionately large in relation to their household income. Nevertheless, the policy of income-related rents on state houses has been a major anti-poverty initiative which initially relieved the considerable financial stress and/or housing instability for more than 54,000 households. As it continues, with some families improving their employment situation and moving out of state housing and as the ongoing building and acquisition programme increases the number of houses, more households will be able to take advantage of it.

Promoting an Employment-Rich Economy and Reducing Unemployment

The Labour-Alliance coalition came into office in 1999 promising active labour market policies that encouraged innovation, higher skill levels and further development of tourism, regional development and the export sector. Most particularly, they wanted job-rich growth. Removing the barriers to employment through education, up-skilling, employment case management, work plans, reducing abatement disincentives, introducing more protective labour relations legislation, and reinvigorating the apprenticeship training scheme were also critical to the strategy. Particular investment was focused on Maori and Pacific employment and capacity-building initiatives

Employment-rich economic development was promoted as being critical to both the country's success in lifting living standards in a modern global market as well as to addressing social equity issues at home. Developing the capacity of the economy to provide greater sustainable employment would enable unemployed people to move off welfare and out of poverty. The aim was to lift living standards, encourage equity and participation, lower welfare costs, increase independence and create more tax payers. This also involved issues concerning the quality of employment, its sustainability, and wage levels. The employment strategy was, among other things, an anti-poverty policy.

The most recent data suggest the strategy has met with some success. GDP growth sits at 4.4%, the highest figure in eight years (SNZ 2003a). (17) For example:

* Unemployment is down to 4.9%, the lowest figure since 1988.

* New Zealand's unemployment rate is lower than that of the United Kingdom (5.1%), the United States (6.0%), Australia (6.2%) and Canada (7.5%).

* Maori unemployment is down from 19.5% in 1998 to 11.4%.

* Pacific unemployment is down from 13.3% in 1998 to 8.2%.

* During the last year the number of jobless people fell by 11,000.

* Total job numbers have increased for the last 10 consecutive quarters (three-monthly periods).

* In the last year there has been a net increase of 44,000 jobs, lifting employment levels by 2.4% (SNZ 2003b).

Interestingly, employment growth has occurred across the whole economy. It has also served to lower welfare recipient numbers, as one would expect. For example:

* Welfare benefit numbers have reduced by over 40,000 since 1999, and by over 62,000 if the sickness and invalid benefits are removed.

* The percentage of the working-age population receiving an income-tested benefit has reduced from 15.9% in 1999 to 14.4% (derived from Work and Income 2003a).

Despite the data, there are still challenges. Although Maori and Pacific unemployment rates have reduced, they are still disproportionately high. Employment conditions were favourable when the Labour-Alliance government took office, with high commodity prices for our primary products. The percentage reduction in benefit numbers is substantial but modest. Nevertheless, the employment and welfare data taken as a whole offer a consistent picture of employment growth and welfare reduction. GDP rates have continued to improve despite the increasing value of the New Zealand dollar, the downturn in international markets since the 11 September attacks in the US and a lower return for our primary products. The consistency is impressive. Its sustainability however, will depend on ongoing job-rich growth through the economic cycles, which would demonstrate a systemic change.

The Primary Health Care Strategy

A quite radical realignment of primary healthcare has only recently begun to take place. Its character and focus are described by the Ministry of Health as:

* a greater emphasis on population health, health promotion and preventative care;

* community involvement;

* involving a range o f professionals and encouraging multidisciplinary approaches to decision making;

* improving accessibility, affordability and appropriateness of services;

* improving co-ordination and continuity of care; and

* providing and funding services according to the population's need as opposed to fee for services when people are unwell (Ministry of Health 2003a).

The strategy was designed in the first instance to overcome the substantial barriers households on lower incomes have when accessing first-contact primary care services, particularly general practitioners. The strategy aims to improve health overall and reduce health inequalities. Over a period of eight to ten years the Government plans to move to a system that provides low-cost access and an increased range of primary healthcare services for all.

As part of the policy to reduce health inequities, funding weightings will favour PHOs where 50% or greater of the enrolled population is Maori or Pacific. The funding formula will also favour PHOs in areas classified Deprivation 9/10, which refers to the New Zealand Deprivation Index's (Crampton et al. 2000) highest areas of deprivation. Implementation has just begun during this current financial year. For this year $50 million has been provided to "Access" fund the first PHOs. Over the next two years a further $360 million will be provided to phase in many more (Ministry of Health 2003b).

This is clearly an anti-poverty health strategy aimed at achieving better health outcomes by, among other things, addressing the unaffordable access to primary health services that many low-income households experience (Waldegrave et al. 1999). To be funded by the Access formula, the PHO will have to commit to low patient fees and an agreed maximum. These will not be set centrally but agreed to between the PHO and the local District Health Board (DHB).

When visiting a general practitioner, no patient will pay above the agreed maximum. These maximums do vary already, however. Children under six will continue to receive free treatment. For the rest of the population, most DHBs and PHOs have agreed to maximums of between $10 and $15. The Auckland DHB, for example, has set a maximum of $10, while Waitemata and Counties have set their maximum at $15. Of concern, some like Rotorua have gone as high as $20, but this is not typical. These are maximums and PHOs are able to charge less. Currently the average cost of a visit to a general practitioner in New Zealand ranges between $45 and $55, with a $15 reduction if the patient possesses a Community Services Card.

It is obviously too early to assess the effectiveness of the new strategy, but it should be noted that unlike most health initiatives in New Zealand, this one has been designed specifically to address the socio-economic determinants of health. Its approach to inequality is quite radical and its design is clearly aimed at increasing access to health services for poor people. Although providing targeted services at first, the strategy will become universal over an eight- to ten-year period.

Income Support to Reduce Child Poverty

As noted earlier, the current Government has committed itself to addressing the rates and thresholds of Family Support payments and the Family Tax Credits. It has also referred to improving annual reviews. These moves have been designed to overcome the systemic problems they inherited, whereby the rates and thresholds declined annually because they were never indexed against inflation. Since 1998, when Family Support was last adjusted, the decline in the real value of the transfer has been 5.5% (MSD 2002). This has saved governments since then around $200 million for Family Support alone.

Unfortunately, these transfers (Family Support and the Family Tax Credits) are the main mechanisms through which the costs for children in low-income houses are funded. It is not surprising there has been a 16% increase in Special Needs Grants in the past two years despite lower benefit numbers and increased employment (MSD 2002). Likewise, the numbers of Special Benefit recipients have leapt 67% and the costs 64% during the last three years (derived from Work and Income 2003c). (18)

Today there are two issues that require addressing as a consequence: there has been a decline in value of these tax and benefit transfers, and they need to be indexed against inflation to ensure the same sort of decline will not occur again in the future.

Action in this area is very important because of the large number of children below the poverty threshold in New Zealand. Throughout the 1990s, the NZPMP consistently reported a third of children to be below the threshold and pointed out that poverty increases with family size. Table 3 earlier in this paper sets out the size of the problem the incoming Government inherited. The Ministry of Social Development reached a similar conclusion in both their Briefing to the Incoming Minister (MSD 2002) and their New Zealand Living Standards 2000 report (Krishnan et al. 2002). The Child Poverty Action Group (CPAG) has repeatedly highlighted the broad dimensions of child poverty across a full range of policy areas and strongly recommended indexing all child payments against inflation (CPAG 2001, 2003). In a major cross-country study Stephens noted the lack of generosity of child payments in the New Zealand welfare system when compared with other like countries (Stephens and Bradshaw 1995, Stephens 2003). Finally, the recent UNICEF report on New Zealand children since the mid-1980s paints a comprehensively grim picture of the poverty impacts on them and their vulnerability (Blaiklock et al. 2002).

Although the Government has indicated it will act on this in its 2004 budget, the extent to which it will address these problems is not yet clear. This initiative will of course be competing with other priorities, like health and education. If the decision is to invest generously, the reduction in the incidence and levels of child poverty will be substantial. The cost to restore Family Support payments to their previous levels will be around $55 million alone. The cost to index them will be on top of that and the Family Tax Credits on top of that again.

Just prior to the 2003 Budget, as this paper was being written, the Minister of Finance stated that if the economy continued to perform well next year, the Government would be able to put an extra $400 million to $500 million into the pockets of poorer families (Dominion Post 8 May 2003). Though not a firm commitment to implement, the statement may be taken as indicative of the Government's intention. This sort of injection of income for poor families with children would substantially reduce current child poverty levels.

Action to restore these tax and benefit transfers to their previous levels will not completely solve the child poverty problem in New Zealand. However, it would substantially decrease child poverty and provide a more secure foundation for children and their families to succeed in education and employment. The indexation would also ensure that the gains in wellbeing would not be lost over time.

THE POLICY CHALLENGES REMAINING

While the five major anti-poverty initiatives assessed above can be expected to achieve a reduction in poverty for many New Zealand households, there are still many challenges remaining. In this section a few brief pointers to some of the critical areas remaining will be outlined.

Housing

Affordable housing is so important for the wellbeing of low-income households. In the budget-setting focus groups that have been continuously run by the NZPMP with a vast array of low-income householders in large urban centres, middle-sized cities and small towns over 10 years, housing consistently appears as the largest budget item. It is usually the first paid item and the quality of other essential goods and services, including food and health care, depends on the residual amount.

The income-related rents policy has been very successful in relieving financial pressures for those fortunate enough to be an HNZC tenant. There are many households in need, however, who are outside that delivery mechanism. As noted earlier, there are currently 11,627 on the state rental waiting list. Of those, 4,587 have been assessed to be in the top two categories of severe or significant persistent housing need (HNZC 2003b). There are also currently 280,706 households receiving the Accommodation Supplement, and therefore judged to have insufficient income to pay their rent/mortgage and other essential costs, none of whom are living in a state rental (Work and Income 2003b). Nearly all of these can be expected to be paying considerably more than 25% of their income on rent or a mortgage. Many in the cities are paying over 40% of their income after receiving the Accommodation Supplement, on housing.

The Accommodation Supplement has become the primary vehicle for state assistance for families unable to meet market housing costs. Too much

pressure is placed on this single blunt instrument and the cost to the state is around $700 million a year (Work and Income 2003c). Its formula in most cases ensures housing costs are well over 30% of income and it is often perceived to be as much a subsidy to landlords as to tenants. Furthermore, it does not help households gather sufficient capital to graduate to a mortgage, (19) leaving many of them vulnerable to a rental market that by international standards is not protective of long-term tenure.

There is a need to develop a broader range of policy instruments, particularly around affordable home ownership and community and non-government partnerships, in a third non-profit sector of rental housing. Home ownership in New Zealand has dropped 6 percentage points in a decade, from 73.8% in 1991 to 67.8% in 2001 (HNZC 2002b). Creative home ownership schemes including rent-to-buy arrangements, mortgage guarantees and top-ups, and HNZC-brokered bulk mortgage agreements with mainstream lenders, could be developed. Home ownership is not a solution for everyone, but many low-income households would prefer to be home owners. Home ownership often provides an accruing asset which changes people's perceptions about themselves in positive and independent ways. It also has the extra advantage of providing freehold ownership in later years when most senior citizens are not part of the work force.

NGO partnerships with HNZC could create an alternative not-for-profit rental sector. Local authorities, iwi and hapu organisations, the churches and larger NGOs are often prepared to share assets and services if they are well supported. These partnerships can save governments large amounts of money and provide well-run, personal housing services. The just released HNZC New Zealand Housing Strategy (2002b) contains a raft of creative policy suggestions from stakeholders, including home ownership options and community NGO HNZC partnerships.

Income Support

Table 3 earlier in this paper sets out for children and adults under 65 years the incidence and severity of poverty to the year 2000. The figures convey the downstream impacts of the poverty-inducing 1991 Budget. In particular, the combination of substantial benefit cuts and market rents on state houses impoverished many households. The failure to address the escalating housing crisis of the 1990s, particularly in Auckland, as rental costs spiralled, further impoverished them.

Remedying the systemic problems of child support transfers, as the Government plans to next year, will certainly relieve many low-income households of much financial pressure, but it will not solve the total poverty problem for many of them. As Table 3 illustrates, the depth of poverty is in marked contrast to that of the superannuitants. The superannuitants themselves are hovering around the poverty threshold. Most are just above it.

Poverty impacts negatively on children's health, their educational opportunities, their long term development and their feelings about themselves and society. There will need to be further income support initiatives for them and their families to overcome the compounding effects of the last decade of under funding. This is particularly appropriate if the goals of participation, equity and an inclusive economy continue to be pursued. Children are vulnerable and they require an equitable start in life that is secure and free from poverty if they are to be expected to develop the resilient skills required for the labour market in a successful modern economy.

The work on New Zealand superannuation offers a model of a comprehensive policy that addresses the issues of equity and sustainability in a transparent and logical manner. The rest of our benefit system is quite inadequate by comparison. The system has historically set levels for payments that have no logical relationship to costs. The budget setting work in the NZPMP indicates they bear little relationship to actual minimal costs households have to carry for food, housing, power, transport, health and so on (Waldegrave et al. 1996). There is a need for greater transparency that relates income support clearly to at least the minimum adequate costs families face.

The work on New Zealand superannuation has achieved that, and it would be clearly unaffordable, and therefore unsustainable, to quickly lift the whole benefit system above the poverty line in the same manner. A goal, however, could be set to lift the value of income support over time, through a mixture of tax and benefit initiatives, to cover the minimum adequate costs a household carries in a transparent manner. A policy goal such as this would require a definition of "minimum adequate costs", budget and living standards research, and an informed public discussion on adequate entitlement. A logical system could be incrementally developed over time which could be expected to be as comprehensive as New Zealand superannuation. This does not necessarily mean it would pay the same amounts, but the same principles of adequacy and sustainability would be addressed. It could be phased in over time in much the same manner as the Primary Health Care Strategy and for the same reasons, those of affordability. In this manner New Zealand superannuation can function as a benchmark for the entire benefit system.

Education, Training and Community Support

Although there has been commendable progress in the training and employment areas as noted earlier in this paper, there are further initiatives that could encourage better synergies between businesses and other enterprises that create jobs, and the training and the educational levels of those who are to be employed, to impact on long-term poverty. Flexibilities and synergies also need to be better developed between social, educational and economic Ministries in the public sector. Likewise, these synergies need to be better developed between these Ministries and the corporate and business sectors on the one hand, and significant community organisations and cultural groups on the other.

Designated individuals in the three parts of the public sector could liaise with the corporate, business and service sectors, identifying where they are investing over the next years and where they would invest if we had the educational and training skills to support it. The information needs to be open and shared across the Ministries, who would together develop a reliable forecast of economic direction and a broad training and educational strategy to support and enrich that direction.

A strategy would then need to be developed that scoped the school leavers, the new entrants to the labour market and people not succeeding in employment to enable the development of appropriate training and education courses that are relevant for those people, and could be expected to prepare them for the actual and anticipated developments in the market. Strategies that will build respect for education, books, creativity, technology and success in communities that have been alienated from mainstream values would also be critical to the process. Cultural loci, particularly the engagement of Maori and Pacific communities sharing in the development of the strategies, would be essential.

A non-silo, holistic, practical approach to lift the skills and educational capacity of communities in line with developments in the economy through high-quality relevant training and education would lift the country's labour market capacity. Catch-up education would need to be a central aspect of the approach. Likewise, openness and cooperation between Ministries and partnerships with significant non-government individuals, organisations and cultural entities in the social and economic sectors would be essential. This is not to suggest that this does not happen at all now, but rather that it does not take place in a comprehensive manner within an overarching strategy.

A balanced and committed strategy of this type that developed a broad base of community and national support could be expected to IlK both the economic and the social capacity of the country, providing more sustainable jobs and reducing the number of benefit recipients.

CONCLUSION

The anti-poverty policies initiated by the two Labour-led coalitions from 1999 onwards are substantial, as the early available evidence set out in this paper indicates. They represent the first official set of major anti-poverty strategies in New Zealand since the early 1970s, and particularly since the country has been fully opened to global markets. The three that have been fully implemented--New Zealand superannuation, income-related rents on state houses and the promotion of an employment-rich economy with reducing unemployment--have been particularly successful in what they set out to achieve. Superannuation is both adequate and sustainable, rents for state houses are affordable, and GDP growth and employment have increased incrementally as unemployment and benefit numbers have decreased.

The Primary Health Care Strategy is an innovative and exciting anti-poverty initiative that is still in its infancy and not yet able to be fully assessed. However, it will certainly increase access to primary health care, reach the most disadvantaged groups and eventually should provide affordable primary health care to all New Zealanders. The planned income support initiatives to reduce child poverty have yet to be spelt out. They obviously cannot be assessed at this stage. Nevertheless, the commitment to address the rates, thresholds and indexing of the key monetary transfers for children next year is the area a majority of researchers and advocates agree to be the most important for any future anti-poverty initiative.

While acknowledging these achievements, there are still pressing challenges ahead if we aim to reduce and eventually eliminate poverty. The focal areas for creative policy development priorities lie in housing for low-income households not in state houses, including home ownership, income support for poor households aged under 65 years, particularly those with children, and a multi-sector approach to the development of relevant education and training that will lift the country's economic and social capacity. The latter will need to be carried out in a culturally informed manner in partnership with Maori and Pacific social policy leaders.

There will always be debate about whether one particular policy like superannuation should be so well resourced when another like housing has only fully addressed one sector of the low-income housing market. Likewise, there will be debate as to whether or not a government should continue to create large fiscal surpluses each year when there are still people below the poverty threshold. These are choices redistributive governments have to make. Those who work in the community, who see much of the pain low-income families experience, want major investments quickly into these social policy areas. Many others in the trade, commerce and servicing sectors, while accepting most aspects of funding public policy and services, seem to become resistant to further investment beyond a certain point.

The Labour-led coalitions have addressed some of the critical areas of poverty in this country. Their approach has been broad-based, involving a mixture of financial assistance, active labour market initiatives and in-kind housing and health transfers. They have developed an integrated redistributive programme which they are delivering in an incremental and measured way. To date they still have the broad base of the New Zealand public behind their policy choices (National Business Review 4 April 2003), which will enable them to continue that policy direction and take up many of the remaining challenges if they choose. Nevertheless, some will argue more could have been achieved or different priorities should have been focused on. That debate is the essence of social policy.

Table 1 Incidence and Severity of Poverty, After Adjusting for Housing
Costs, 1993-1998

Poverty Measure      Poverty Incidence

                   Household    Individual

60% 1993             18.5%         20.5
60% 1998             19.3%         20.6

Poverty Measure       Poverty Reduction

                   Incidence    Poverty gap

60% 1993             58.1          51.3
60% 1998             42.9          80.5

Poverty Measure      Poverty          Poverty Gap

                   ($ millions)     (% poverty line)

60% 1993             826.45               29.7
60% 1998           1,010.00               28.8

Source: Calculated from NZPMP data base.

Table 2 Incidence and Severity of Poverty, by Housing Tenure, 1998

Tenure             Incidence (%)       Structure(%)
                                   Before Adjusting for
                                      Housing Costs
Owned
  With mortgage         8.4                17.2
  No mortgage          15.0                38.9
Rented
  HousingNZ            36.9                13.4
  Employer             14.9                 2.2
  Private              18.1                17.9
All Tenure             15.4               100.0

                                   After Adjusting for
                                      Housing Costs

Owned                  17.2                28.0
  With mortgage         4.7                 9.8
  No mortgage
Rented                 71.9                20.7
  HousingNZ            17.3                 2.1
  Employer             32.2                25.4
  Private              48.2                14.0
All Tenure             19.3               100.0

Tenure             Poverty Gap ($m)

Owned
  With mortgage          69.0
  No mortgage           139.7
Rented
  HousingNZ              66.5
  Employer                3.5
  Private                81.2
All Tenure              400.7

Owned                   245.0
  With mortgage          83.2
  No mortgage
Rented                  254.9
  HousingNZ              18.7
  Employer              280.3
  Private               128.0
All Tenure            1,010.1

Source: Derived from the NZPMP database.

Table 3 Incidence and Severity of Poverty, 60% of Median Equivalent
Household Disposable Income, People, Adults and Children, 1997 &
2000 *

(a) Before adjusting for housing costs

Poverty                  All                 Adults 18-64
Incidence          1997       2000          1997       2000

Market             28.4       27.4          18.2       18.3
Disposable         15.7       16.3          11.0       13.2
Efficiency         44.7%      40.5%         39.6%      27.9%

Poverty Gap $m
Market            5,668      6,625           993       1,555
Disposable          598        729           235         341
Efficiency         89.4%      89.0%         76.3%      78.1%

(b) After adjusting for housing costs

Poverty                  All                 Adults 18-64
Incidence          1997       2000          1997       2000

Market             29.7       30.2          19.5       21.2
Disposable         20.3       21.9          16.7       18.8
Efficiency         31.6%      27.5%         14.4%      11.3%

Poverty Gap $m
Market            6,059      7,283          1,059      1,654
Disposable        1,211      1,589            422        626
Efficiency         80.0%      78.2%         60.2%      62.2%

(a) Before adjusting for housing costs

Poverty              Adults 65+             Children 0-18
Incidence          1997      2000           1997      2000

Market             76.9       71.2          30.8       29.9
Disposable         30.0       17.9          20.5       23.9
Efficiency         61.0%      74.9%         33.4%      20.1%

Poverty Gap $m
Market            3,226      3,440          1,449      1,630
Disposable          117         34            246        354
Efficiency         96.4%      99.0%         83.0%      86.3%

(b) After adjusting for housing costs

Poverty                Adults 65+          Children   0-18
Incidence           1997      2000          1997      2000

Market             75.3       70.6          33.5       33.9
Disposable         12.3       10.8          33.5       35.0
Efficiency         83.7%      84.7%          0.0%      -3.2%

Poverty Gap $m
Market            3,222      3,438          1,778      2,191
Disposable          146        115            643        848
Efficiency         95.5%      96.7%         63.8%      61.3%

Source: NZPMP database.

* These dates would normally be written up as 1998 and 2001 because
they are derived from the 1997/98 and 2000/01 HES years. The actual
months involved were 1 April 1997 to 31 March 1998 and 1 July 2000 to
30 June 2001, respectively. Many of the questions and the diary keeping
involved 12 months' recall to give annual amounts. Our advice from
Statistics NZ is to centre the survey. In other words, the real bases
for these two surveys are September 1997 and December 2000. This means
the 2000/01 HES is not able to provide helpful data concerning the
policy changes referred to in this paper apart from New Zealand
superannuation, which began on 1 April 2000. The income-related rents
were introduced 1 December 2000, too late to be picked up adequately
in 2000/01 HES database. To make that clear we have, in this Table
only, referred to the 1997/98 year as 1997 and the 2000/01 year as
2000.

Note: "Market" = market income only; "Disposable" = net personal income
after taxation has been deducted and cash assistance from government
has been added; "Efficiency" = the effectiveness of the tax and benefit
system to reduce poverty

(2) The evidence is restricted to the available data at the time of writing early in 2003. In later years when the Household Economic Survey (HES) database and the Ministry of Social Development's Economic Living Standards Index (ELSI) (Krishnan et al. 2002) can be applied to the full implementation of the new social and economic policies, a more robust assessment will be able to be made.

(3) The debate is about more than semantics. The approach adopted not only affects to varying degrees the reported incidence of poverty at a population and sub-group level, but also leads to there being significantly different sets of citizens identified as poor. See Perry (2002) for a recent review of the "mismatch" and related conceptual issues.

(4) This research was funded by the Foundation for Research, Science and Technology. The NZPMP was undertaken by three organisations: Business Economic Research Limited (BERL), The Public Policy Group at Victoria University of Wellington (VUW) and the Family Centre Social Policy Research Unit (FCSPRU). The research leaders of the project were Paul Frater (BERL), Robert Stephens (VUW) and Charles Waldegrave (FCSPRU). Currently the project continues under the leadership of Robert Stephens and Charles Waldegrave.

(5) The 60% threshold is not fixed. Substantial policy changes or a major economic shift could be expected to lead to different focus group results for the level of a minimum adequate income and a consequent different percentage of the median.

(6) The focus groups indicated that costs in Auckland were a little above the line and those in rural areas somewhat below it, but these distinctions could not be shown in the Household Economic Survey (HES) data.

(7) These data are cross-sectional for each year. It is not a longitudinal study that could measure the dynamics of those in poverty and inform us of how many different people move in and out of poverty each year.

(8) The random sample consisted of 401 people selected from throughout the country. The margin of error was plus or minus 4.9 percentage points at the 95% confidence level.

(9) It should be noted, though, that some adjustments were made to benefit supplements during a period of economic growth in the mid-1990s, which led to a marginal reduction in poverty at that time. Furthermore, visits of children under six years to a general practitioner became free of charge nearer the end of the period.

(10) They did not actually reduce the income of those already receiving superannuation. Instead, they linked the annual increases to prices rather than wages, with the planned effect of reducing payments to 60% of the average wage within five years.

(11) The living standards of older New Zealanders research showed that the elderly renting Housing NZ houses actually paid more than those in the private sector (Fergusson et al. 2001:51). The HES database showed that the proportion of households on the lowest incomes (bottom 20%) who were paying 30% or more of their income in rent increased from 28% in 1991 to 44% in 1998 (Mowbray 2001:42).

(12) The Accommodation Supplement thresholds have likewise not been adjusted, compounding the problem.

(13) There has been some discussion by researchers and advocates about whether New Zealand Superannuation should have been addressed prior to the needs of poor households with children. The order of policy change is certainly debatable. The Government argues that many low-income families have been helped directly by the income-related rents policy, the active labour market policies and the public health strategy. Furthermore, they say most younger families still have the opportunity to improve their situation, whereas most of the elderly have completed their working years and do not have that possibility. The opposing argument states that although income for the elderly had started to deteriorate, it was still well above that for poor families and it would have been wiser to begin with them. Prioritising the needs of one group over another is always very difficult. From the perspective of these authors, the overall balance of policy mix is very important and that is addressed in the five policies highlighted in this paper. Unfortunately some have to take priority, and arresting the fall in the value of superannuation restored the only government social assistance payment that was previously above the poverty line. That has preserved a benchmark in terms of adequacy and sustainability, which we argue later could usefully be applied to the whole benefit system.

(14) This was presumably because the Act also legislated for the New Zealand Superannuation Fund, which required more time to develop.

(15) The process of standardizing HES to the census population rather than using the HES sample resulted in a small rise in median earnings and a consequential rise in the measured poverty incidence from 20% in 1997 (Stephens and Waldegrave 2001) to 30%, as set out here. The average poverty gap, though, was only a few dollars.

(16) It should be noted that the introduction of income-related rents was accompanied by a new social allocation methodology that better targeted needy people instead of simply relying on who could afford to pay the previously high rents on HNZ properties.

(17) Some caution needs to be taken concerning future projections given the impacts of the war in Iraq, the Severe Acute Respiratory Syndrome (SARS) virus and the high level of the New Zealand dollar.

(18) It is worth noting here that the research and advocacy of the Wellington Downtown Community Ministry group has contributed to the increased awareness and uptake of this supplementary benefit (Howell et al. 2000). This increased uptake is also probably affected by the non-indexation of the Accommodation Supplement cap.

(19) Although the Accommodation Supplement can also be paid to home owners with a mortgage, it is insufficient by itself, because of its formula, to help people save for a deposit.

REFERENCES

Birch, Hon W (1998) "Budget policy statement" 9 December, The Treasury, Wellington.

Blaiklock, A., C. Kiro, M. Belgrave, W. Low, E. Davenport and I. Hassall (2002) "When the invisible hand rocks the cradle: New Zealand children in a time of change" Innocenti Working Papers 93, UNICEF Innocenti Research Centre, Florence.

Cartwright, Dame S. (2002) "Speech from the throne" 27 August, http://www.beehive.govt.nz/throne.cfm

Child Poverty Action (2001) Family Financial Assistance 1986-2002 Backgrounder, revised, Child Poverty Action, Auckland.

Child Poverty Action Group (2003) "Submission Relating to the Budget Policy Statement" 28 January, Child Poverty Action Group, Auckland.

Clark, Rt. Hon. Helen (2002) "Prime Minister's speech to address-in-reply debate" 28 August, http://www.beehive.govt.nz/ViewDocument.cfm?DocumentID=14762

Crampton, P., C. Salmond and J. Kirkpatrick (2000) Degrees of Deprivation in New Zealand: An Atlas of Socioeconomic Difference, David Batemen, Wellington.

Cullen, Hon. M. (2002) "Budget speech" Treasury, Wellington, http://www.treasury.govt.nz/ budget2002/speech/#top

Department of Labour (2003) Employment Strategy: Progress to Date July 2001-December 2002, Department of Labour, Wellington.

Department of Social Security (1993) Households Below Average Incomes: a Statistical Analysis, 1979-1990/1, HMSO, London.

Fergusson, D., B. Hong, J. Horwood, J. Jensen and P. Travers (2001) Living Standards of Older New Zealanders, Ministry of Social Development, Wellington.

Gosche, Hon. M. (2000) Statistics, Office of the Minister of Housing, Wellington.

HNZC (2002a) HNZC 2001/02 Annual Report, Housing New Zealand Corporation, Wellington.

HNZC (2002b) New Zealand Housing Strategy: Reports from Stakeholders, Housing New Zealand Corporation, Wellington.

HNZC (2003a) HNZC Asset Management Plan 2002/2003, Housing New Zealand Corporation, Wellington.

HNZC (2003b) "Latest Statistics as at March 2003" Housing New Zealand Corporation, Wellington.

HNZC (2003c) "Rental Homes, Step 5" http://www.hnzc.govt.nz/rental/step5.htm

Howell, G., D. Simmers and G. Hackwell (2000) Still Missing Out, Downtown Community Ministry, Wellington.

King, Hon. Annette (2002) "Implementing the Primary Health Care Strategy and improving access to primary health care services" Memorandum to Social Policy and Health Committee, 2 March, www.executive.govt.nz/minister/king/cabinet02-03/02.htm

Krishnan, V., J. Jensen and S. Ballantyne (2002) New Zealand Living Standards 2000, Centre for Social Research and Evaluation, Ministry of Social Development, Wellington.

Louisson, Simon (1991) "N.Z. budget slashes into welfare state, cuts spending" Reuters News 30 July.

McCulloch, B. and J. Frances (2001) "Financing New Zealand Superannuation" Treasury Working Paper 01/20, The Treasury, Wellington.

Ministry of Health (2003a) "Primary Health Care Strategy" website home page, http:// www.moh.govt.nz/primaryhealthcare

Ministry of Health (2003b) "Primary Health Care Strategy Implementation--Q&As" http:// www.moh.govt.nz/moh.nsf/wpg_Index/About-Primary+Health+Care+Strategy+ Implementation

Ministry of Social Development (2000) "NZ Superannuation Weekly Rates: Benefit Rates on or after 1 April 2000" Ministry of Social Development, Wellington.

Ministry of Social Development (2002) Briefing to the Incoming Minister, Ministry of Social Development, Wellington.

Ministry of Social Policy (1998) "Unpublished proposal to Ministers" Ministry of Social Policy, Wellington.

Mowbray, M. (2001) Distributions and Disparity: New Zealand Household Incomes, Ministry of Social Policy, Wellington.

Perry, B. (2002) "The mismatch between income measures and direct outcome measures of poverty" Social Policy Journal of New Zealand, Issue 19.

SNZ (2001a) Demographic Trends 2001, Statistics New Zealand, Wellington.

SNZ (2001b) "New Zealand Superannuation receivers for the year ended June 2001" Household Economic Survey, Statistics New Zealand, Wellington.

SNZ (2003a) "Gross Domestic Product, December quarter, 2002" Statistics New Zealand, Wellington.

SNZ (2003b) "Household Labour Force Survey December quarter 2002" Statistics New Zealand, Wellington.

SNZ (2003c) "National population estimates, December quarter, 2002" Statistics New Zealand, Wellington.

Stephens, R. (2003) "The level of financial assistance to families with dependent children: a comparative analysis" Social Policy Journal of New Zealand, Issue 20, July.

Stephens, R. and J. Bradshaw (1995) "The generosity of New Zealand's assistance to families" Social Policy Journal of New Zealand, Issue 4.

Stephens, R. and C. Waldegrave C (2001) "The effectiveness of the transfer and tax system in reducing poverty in 1998" Social Policy Journal of New Zealand, Issue 16.

Stephens, R., C. Waldegrave and P. Frater (1995) "Measuring poverty in New Zealand" Social Policy Journal of New Zealand, Issue 5.

Stephens, R., C. Waldegrave and P. Frater (2000) "Below the line: an analysis of income poverty in New Zealand, 1984-1998" Graduate School of Business and Government Management Working Paper 2/00, Victoria University of Wellington.

Treasury (2000) Budget Economic and Fiscal Update, Treasury, Wellington.

Waldegrave, C. (2000) "Assessing the impacts of the New Zealand housing reforms during the 1990s" Ministry of Social Policy Seminar Series, 25 October, www.msd.govt.nz/ publications/docs/charleswaldegrave.pdf

Waldegrave, C., P. Frater and R. Stephens (1997) "An overview of research on poverty in New Zealand" New Zealand Sociology, 12(2).

Waldegrave, C., P. King and S. Stuart (1999) The Monetary Constraints and Consumer Behaviour in New Zealand Low Income Households, The Family Centre Social Policy Research Unit, Lower Hutt, Wellington.

Waldegrave, C. and R. Stephens (2000) "Poverty: the litmus test of social and economic policy failure" Proceedings of the Second Biennial Aotearoa New Zealand International Development Studies Network (DEVNET) Conference, 17-19 November 2000, Victoria University of Wellington, DEVNET, Wellington www.devnet.org.nz/conf/Papers/ waldegrave.pdf

Waldegrave, C., S. Stuart and R. Stephens (1996) "Participation in poverty research: drawing on the knowledge of low-income householders to establish an appropriate measure for monitoring social policy impacts" Social Policy Journal of New Zealand, Issue 7.

Work and Income (2003a) "Numbers on benefits, March 2003" Ministry of Social Development, Wellington.

Work and Income (2003b) "Quarterly client profile, December 2002" Ministry of Social Development, Wellington.

Work and Income (2003c) "Unpublished time series data" Ministry of Social Development, Wellington.

Charles Waldegrave (1)

Robert Stephens

Peter King

New Zealand Poverty Measurement Project

(1) Acknowledgements

NZPMP gratefully acknowledges the support of the H.B. Williams Turanga Trust for helping resource the research for this paper. The ongoing funding for the NZPMP by the Foundation for Research, Science and Technology is also gratefully acknowledged and referred to later in this paper.

Correspondence

Charles Waldegrave, The Family Centre Social Policy, Research Unit, PO Box 31050, Lower Hutt, Wellington, e-mail [e-mail]; Robert Stephens, The School of Government, Victoria University of Wellington, PO Box 600, Wellington, e-mail [e-mail]; Peter King, The Family Centre Social Policy Research Unit, PO Box 31050, Lower Hutt, Wellington, e-marl [e-mail]