About DPMC

10. Impact on costs and tenure choice

• All measures of housing affordability have declined in the past five years.
• Rents have increased in line with incomes, however, there are emerging issues about rental affordability among low-income groups.
• Home ownership rates have fallen and are projected to fall further in the next 10 years.
• There is a growing group of households that cannot afford to buy a home at current prices, interest rates and income, and are not eligible for assistance through the Housing New Zealand Corporation.

10.1 A definition of housing stress

Yates and Milligan (2007), in a report for the Australian Housing and Urban Research Institute, define a benchmark of housing costs of 30% or 40% of income as a measure of housing stress. The 2007 Social Report (Ministry of Social Development) shows that in 2004, 22% of New Zealand households spent more than 30% of their net income on housing costs. This represents a decline from 24% in 2001, however, since the late 1980s there has been a substantial increase in the proportion of households spending more than 30% of their income on housing. By age group the proportion of people living in households spending more than 30% of the income on housing is concentrated largely among people under the age of 45, with the highest shares in people under 18 (29.2%) and people aged 18 to 24 (29.0%).

Figure 27: Proportion of households spending over 30% of net income on housing

Source: 2007 Social Report (Ministry of Social Development)

Some households will have made a conscious decision to spend more than 30% of their income on housing and will feel that they can afford to do so. The biggest concern is amongst low-income earners because they have less residual income to meet the cost of other living expenses. The proportion of households in the lowest 20% of the equivalised household income distribution spending more than 30% of their income on housing rose from 16% in 1988 to a peak of 49% in 1994 and fell to 35% by 2004.

10.2 Impact on tenure type – home ownership affordability has declined

Home ownership affordability is highly cyclical, moving in line with changes in house prices, interest rates and income.

There are a variety of measures of housing affordability commonly reported in the media. Massey University produce the National Home Ownership Affordability Index using data on median house prices, average earnings and interest rates. The Fairfax Media Home Loan Affordability Index calculates a measure of disposable income, based on the amount of income a buyer has left after paying the mortgage. In addition, the economics departments of a number of banks calculate and publish affordability measures. The National Bank publishes a measure of affordability that calculates the proportion of household weekly earnings required to service a 25-year mortgage based on a two-year fixed rate and a 20% deposit (National Bank of New Zealand, 2007). This measure shows that the proportion of income required fell from around 40% for most of the period from 1992 to 2000 to around 35% between 2000 and 2003 and has subsequently increased to 60%. Figure 26 in section 8 shows that house price to income ratios have lifted well above their average. Analysis of regional house price to income ratios shows that prices are most stretched in Auckland.

Measures of affordability generally attract a degree of controversy, with much debate about the appropriateness of the measure, what it includes and who it is applicable to. House price to income ratios are often criticised for not capturing the impact of interest rate changes.

The Unit has developed two additional measures of affordability. The first provides a time series of the number of hours that a purchaser needs to work to pay off their mortgage. The second uses Survey of Family Income and Employment data to calculate the number of people who can afford to service a mortgage at existing prices, income and interest rates.

Measure one: a real affordability index

Figures 28 and 29 shows how affordability indices based on real and nominal interest rates have evolved in the last two decades. The real interest rate is calculated by deflating the nominal interest rate by the average of the previous four quarters’ and the forthcoming four quarters’ change in the CPI.[20] The formulae for the two indices are:

Nominal 		=      nominal interest rate  x Quotable Value house price index
affordability index      			average hourly earnings

Real  			=      real interest rate  x Quotable Value house price index
affordability index      			average hourly earnings

In figure 28, the indices are rebased so the Quotable Value price index is equal to $295,000 in March 2006, the median house price in that quarter. The indices have the interpretation of the number of hours of work needed to pay the annual nominal or real interest cost of a mortgage just large enough to purchase a median priced house, including principal payments. In figure 28 the indices are rebased so that they both have an average of 1,000 for the period March 1992–March 2007. It should be noted that a rise in the index means that housing is less affordable to those contemplating borrowing to purchase a house.

This measure of affordability does have some limitations. Average hourly earnings do not include all sources of income and are affected by changes in the composition of employment, even if there are no changes in actual hourly wages. In addition, they relate to a particular job rather than the income of a household. While the measure does have limitations, it adds to the suite of affordability measures that can be considered.

Figure 28: Indicator of real and nominal affordability

Source: RBNZ

Figure 29: Indicator of real and nominal affordability relative to historical average

Source: RBNZ

• The nominal affordability index reaches its highest level over the entire period in 2007, 45% above its average level. In contrast, the real affordability index was 30% above its average value in March 2007.
• Nominal and real affordability are volatile series.
• While both real and nominal housing affordability have deteriorated sharply since March 2002, both the nominal and real housing affordability indices were at very low levels between 1999 and 2001 – housing was relatively affordable. Home ownership declined between 1996 and 2001, despite relatively good affordability towards the end of the period, suggesting factors other than affordability affected ownership rates over this period, or that poor affordability from 1996-1999 had a large impact on ownership rates.
• From figures 28 and 29, it is apparent that the number of hours required to service the nominal interest cost of a mortgage is considerably larger than the number of hours required to service the real interest cost of a mortgage. It follows that part of the deterioration in nominal housing affordability in the past five years reflects the higher average inflation rate.

Measure two: the number of people who can afford to service a mortgage

The Unit has used Survey of Family Income and Employment (SOFIE) data to analyse how the number of couples and non-partnered individuals that can afford to buy a home has changed over time. In all of this analysis couples and non-partnered individuals include people with children and people without children, consistent with the definitions used in SOFIE. For example, the category called couples would include a young couple with no children and a married couple with three teenage children.

The income data discussed here is based on 2004 SOFIE results, with income indexed forwards to 2006 and back to 2000 based on changes in income[21]. The approach is to estimate the proportion of renting couples and non-partnered individuals that can afford a lower-quartile-price house in their region and keep their mortgage payments, including principal reductions, at a maximum of 30% of their gross income.[22]

In 2006, 29% of renting couples (including couples with and without children) and 2% of renting non-partnered individuals (including individuals with and without children) could afford to buy a lower-quartile house in their region.[23] Based on the interest rates and house prices that prevailed in 2000, with income indexed back to 2000, 59% of couples and 11% of non-partnered individuals could have afforded a lower-quartile-price house. These numbers suggest a large decline in affordability between 2000 and 2006. This is shown graphically in figure 30.

Figure 31 provides the equivalent numbers if the calculation excludes renters who have net worth in excess of 20% of the price of a lower-quartile-price house.[24] This excludes a group who may have voluntarily decided not to use their net worth as equity to buy a house. Of course, even if this group did want to use their net worth to buy a house, not all those in it would have the income to service a mortgage and it may not be reasonable to expect people to convert some types of net wealth into a home deposit.

Figure 30: Percentage of renters that can afford a lower-quartile-priced house	Figure 31: Percentage of renters that can afford a lower-quartile-priced house, excluding those with net worth 20% of price
Source: SOFIE, Treasury	Source: SOFIE, Treasury

At current prices and interest rates, the gap between the income of those who cannot afford to buy a lower-quartile-priced house and the income needed to service a mortgage on a lower- quartile-priced house while keeping the payments at or below 30% of income is large. On average, to bridge the gap, a couple would need a deposit of $122,000, and non-partnered individuals would need a deposit of around $170,000. For many couples and individuals saving a deposit this large is likely to pose a considerable hurdle to home ownership, except perhaps for those at the upper end of the income distribution. Eliminating student debt would make no discernable difference to the overall numbers because the median deposit gap is too large to be offset by student debt.[25]

According to this measure, by region, affordability is most stretched in Auckland. Of the 71% of couples that are estimated to be unable to afford a lower-quartile-priced house, close to half of these couples are in Auckland. Only 6% of renting couples in Auckland could afford a lower-quartile-priced house, compared with 29% nationally.

This measure can be used to consider a number of different groups within the population. For example, affordability can be calculated based on the number of children that non-partnered individuals and couples have. The numbers are broadly the same as those discussed above and are shown in table 12. In the table each of the non-partnered individuals and couples groupings are separated into income quartiles and into three different groupings based on the number of children they have. For example, 20% of non-partnered individuals in the fourth income quartile with one child can afford a lower quartile house in their region.

Table 12: Affordability by number of children and income

	Percentage who can afford lower quartile house by number of children
		One child	Two + children	No children	Total
Non-partnered individuals
Income quartile
1  (Less than $14,803)		0	0	0	0
2  ($14,804 to $23,463)		0	0	0	0
3  ($23,464 to $39,665)		0	5	10	10
4  ($39,666 and above)		20	1	2	2
Total		4	1	2	2
Couples
Income quartile
1  (Less than $41,175)		0	0	0	0
2  ($41,176 to $66,955)		0	1	1	1
3  ($66,956 to $100,246)		40	39	39	39
4  ($100,247 and above)		88	89	89	89
Total		27	30	29	29

 

Source: SOFIE, The Treasury

Home ownership affordability: conclusion

Affordability is a function of house prices, interest rates, income and initial wealth, or the equity that a household has available. A wide range of measures of affordability show that affordability has declined in the past four to five years. As a result, there is a growing group, sometimes described as the ‘intermediate market’, that cannot afford to service a mortgage at prevailing prices, interest rates and income, and are not eligible for state assistance through Housing New Zealand Corporation.

10.3 Impact on tenure type – home ownership rates have fallen and are projected to continue falling

Ownership rates increased from around 57% in 1945 to just over 71% by 1981. As shown in table 2 in section 4.2, between 1981 and 1991 home ownership rates increased by around 3 percentage points. Since 1991, home ownership rates have declined, reaching 66.9% in the 2006 Census. Around 28% of households rent from private landlords.

An international comparison

In their May 2007 submission to the Select Committee of Inquiry into Housing Affordability in New Zealand, Westpac provided the following international comparison of home ownership rates.

Table 13: Home ownership rates (percentage)

	1960-1970	1980-1990	200-2003
NZ	68	72	67
Australia	65	69	71
Canada	53	61	67
France	43	53	57
Germany	33	39	40
Ireland	66	78	80
UK	36	63	68
USA	64	66	68

 

Data for 2006 shows that home ownership rates in the United Kingdom have further increased to just over 70%. New Zealand’s experience has been somewhat different to the other countries in table 13. Most of these countries have had a large increase in house prices but have generally had increasing home ownership rates, in some cases facilitated by significant subsidisation by government, for example, the Right-to-Buy programme for state housing tenants in the United Kingdom. While aggregate home ownership rates in the other countries in table 13 have either increased or remained broadly constant, there are declining home ownership rates in New Zealand.

Changes by region

Between 1986 and 2006, Auckland, Nelson and Bay of Plenty experienced the largest percentage point falls in home ownership. Between 2001 and 2006, the largest percentage point falls in home ownership have been in West Coast, Taranaki and Waikato. Home ownership rates are highest in Tasman and Southland, and lowest in Auckland.

Table 14: Regional profile of home ownership

Percent of homes owner-occupied	1986	1991	1996	2001	2006
Northland	72.3	73.9	71.2	70.5	68.6
Auckland	74	72.7	69.2	64.6	63.8
Waikato	70.2	71.4	68	67.8	65.4
Bay of Plenty	74.4	76.2	71.7	68.4	67.3
Gisbourne	67.7	67.2	65.1	63.2	61.8
Hawkes Bay	73.1	73.7	70.6	67.8	67.9
Taranaki	73.9	75.1	72.1	72.2	69.9
Manawatu-Wanganui	70.8	71.7	68.7	67.9	66.8
Wellington	71.9	72.1	69.9	66.9	66.1
Tasman	77.1	78.3	74.7	73.7	72.5
Nelson	79.6	77.3	72.3	68.7	68.6
Marlborough	77.3	80.1	77.4	76.1	75.8
West Coast	74.2	76.2	73.7	72.6	69.3
Canterbury	76.7	76.6	73.8	71.4	70.4
Otago	74.8	74.7	71.9	69.6	69.1
Southland	79.3	80.3	77.7	75.6	73.5
New Zealand	73.7	73.8	70.7	67.8	66.9

 

Source: DTZ 2007

Changes by age
The decline in home ownership has been larger for younger households than for older households. The largest falls in home ownership have been in the 25-29 and 30-34 age groups. Between 1986 and 2006, home ownership rates fell 17.9 percentage points in the 25–29 age group and 17.7 percentage points in the 30–34 age group. The next largest falls were in the 35–39 (15.5 percentage points) and 40–44 (12.2 percentage points) year age groups. For most age groups, the largest falls have come in the past 10 years, with the biggest falls in the group that historically make up first home buyers. For each of the most affected groups discussed above, there was a fall in home ownership rates of around 5 percentage points between 1996 and 2001 and a further fall of 3–5 percentage points between 2001 and 2006.

Analysis of different age cohorts suggests that the time it has taken successive household cohorts to enter home ownership has lengthened. In addition, analysis by Morrison (2007, forthcoming) shows that deferral of home ownership only partly explains declining home ownership, with a structural downward shift in home ownership rates also being an important factor.

Figure 32: Home ownership by age

Source: DTZ, 2007, derived from Census

Changes by ethnicity, income and sex
In line with the general decline in home ownership, Centre for Housing Research in Aotearoa New Zealand (CHRANZ) reports declines in home ownership across all ethnic groups, with the biggest percentage point falls in the Asian ethnicity group.[26]

Table 15: Home ownership by ethnicity

	1986	1991	1996	2001	2006
Asia	69.1	62.7	61.9	58.8	56.5
European	76.4	77.1	74.1	71.9	70.5
Maori	49.2	52	48	44	42.5
Not elsewhere included	69.1	64.7	59.4	58.5	53.5
Other	56.3	52.4	39.7	32.8	42.3
Pacific Peoples	44.5	43.7	40.2	35.5	34.1
Middle Eastern/Latin Am/African					33.2
New Zealander					76.1

 

 

Source: CHRANZ

Rates of home ownership have also tended to fall by household income group. CHRANZ (2007) show declining home ownership rates across income groups since 1986. Not surprisingly, in 2006 the highest rates of home ownership in every region in New Zealand were in the $100,000 plus household income group. The strong real income growth results, discussed in section 4.5, amongst the highest parts of the income distribution have contributed to high home ownership rates in this group.

Home ownership patterns also have an important gender dimension. Older women have high rates of home ownership, largely due to their tendency to outlive their spouses. Dwyer (2007) notes that in the 2006 Census, slightly more women than men owned or partly owned the dwelling they lived in, partly due to the older age structure of the female population. Sole parent households, which have lower home ownership rates than most other household types, tend to be headed by women. The income level of women is lower than men (Statistics New Zealand, 2007), which may also affect home ownership rates, particularly for sole parent households.

Future outlook for home ownership

Projections in research commissioned by the Centre for Housing Research in Aotearoa New Zealand (DTZ, 2007) suggest that home ownership rates are likely to decline further in the future. Based on demographic and population projections, and expectations of high prices, DTZ project that the home ownership rate will fall a further 5 percentage points by 2016 to 62%. To avoid this fall in the home ownership rate, around 10,000 additional renting households would need to move into home ownership each year.

10.4 Impact on tenure type – rental affordability

Despite small changes in rents relative to house prices, housing costs for the lowest 20% of income earners have increased sharply over the past 20 years, and there are signs of increasing housing stress amongst the lowest income earners, particularly beneficiaries (Ministry of Social Development, 2007). Benefits are indexed to the CPI, so if rents increase by more than the CPI, beneficiaries will be negatively affected. Part of the increase in housing stress may also reflect the Accommodation Supplement maxima, which are not indexed and were last changed in 2005 when they were set at 2003 rent prices. The number of renters receiving the maximum Accommodation Supplement has increased from around 40,000 in 2005 to 60,000 in 2007.

The rising housing stress for the lowest 20% of income earners also reflects changes in income more widely, as discussed in section 4.5, with little real income increase between 1982 and 2004 in the lowest parts of the income distribution. Proportionately, housing stress is greater for renting households than for owner-occupied households with similar incomes. In 2004, 35% of renting households spent more than 30% of their income on housing, compared with 15% of home owners. The different age structures of the groups makes this comparison complicated as many older people own their home and have different expenditure and income from younger groups. CHRANZ (2004) reports that in Auckland renting households with the highest levels of stress are in the lower to middle-income bands, have children in the household, have five or more people living in the dwelling and either unemployed or not actively seeking work as well as not being retired.

Figure 33: People receiving maximum Accommodation Supplement	Figure 34: Real income by percentile
Source: Ministry of Social Development	Source: Ministry of Social Development

While rental affordability is not a widespread problem at present, rents could increase in the future, particularly as capital gains diminish and yields are rebuilt through increasing rents or private sector investment in new housing supply diminishes.