About DPMC

7. Supply factors

• The supply of new occupied dwellings has responded strongly to population growth, however, there are some signs of a shortage of supply in Auckland, particularly Manukau.
• The increase in new supply of occupied dwellings has come with large increases in the cost of constructing new dwellings since 2001, including increases in section prices, labour costs and the cost of materials.
• New dwellings are only a relatively small share of house sales.

7.1 Supply of dwellings

Magnitude of the increase in dwellings

New Zealand does not have an annual survey of the size of the country’s dwelling stock. The Census provides a snapshot of the number of dwellings every five years. In between these years, Statistics New Zealand estimates the size of the dwelling stock based on the number of building consents that are approved, incorporating some assumptions about the number of consents that are not acted on and the number of deletions from the housing stock.

Housing supply can be slow in responding to changes in population because of the time taken to assemble the required mix of land, materials and labour; planning delays; and the time required to construct a dwelling. Growth in building consents and dwellings lagged behind the sharp population growth associated with net migration inflows during 2002 and 2003, however, by June 2003 consents for new dwellings reached 30,000 per annum and increased further to more than 33,000 in the year to June 2004, before dropping back to around 26,000 a year in the 2005, 2006 and 2007 June years. This represented a sizable step up in building activity from a historical average of around 20,000 building consents a year, and it shows a construction sector that is able to respond to rising demand and prices.

The net result of this strong increase in building activity was an increase in dwellings between the 2001 and 2006 Censuses of 125,000 dwellings. Around 110,000 of these were new occupied dwellings, with an increase of around 15,000 unoccupied dwellings. This large increase in dwellings is somewhat different to the experience of some other countries that have experienced house price booms, particularly the United Kingdom, where the supply response has been muted.

The critical issue examined by the House Prices Unit is whether the increase in supply has been sufficient to match the increase in the population. If not, then a shortage of houses would be expected to flow into higher house prices, higher rents and signs of crowding as more people are forced to live in a limited number of houses. This section also includes a discussion of the type of new dwellings that have been built.

A simple comparison of the growth in the total number of dwellings (8.1%) between the 2001 and 2006 Censuses compared with population growth (7.7%) does not point to any shortages at the aggregate level. A comparison between the growth in occupied dwellings (8.1%) and population and household growth does also not point to any shortages at the aggregate level.[11] This simple comparison could be misleading, for example, the regional distribution of dwelling and population growth is important, while the distribution of growth within the five year period could also be important.

The Unit has investigated two further approaches to address the question of whether the increase in supply has been enough to house the population growth. It is also important to consider the type of supply that has been added and which segments of the population have accessed the new supply.

Approach 1: Comparison of building consents and population growth

The first approach is to estimate the number of dwellings required to house the growth in the population, building in the following two assumptions:

• In the 2001 Census, the average household size was 2.6 per house. In the 2006 Census, the average household size was 2.7 people per house. The table calculates the number of houses required if average household size is 2.7 and if the ‘desired’ size was actually 2.5, in line with a declining trend over time.
• Building consent data is used to estimate the rate of construction of new houses, with an assumption that only 80% of new building consents translate into net additions to the country’s occupied housing stock. The remaining 20% is accounted for by consents not acted on, holiday homes and by activity that is replacing deletions from the housing stock.

The table below shows the assessment of undersupply and oversupply resulting from this analysis. This provides an alternative approach to the simple comparison of Census numbers and sheds some light on the pattern of building between Census years.

The period from 1995–1997, also a period of large population growth, shows signs of an undersupply of houses. This was subsequently filled by a strong period of building activity between 1998 and 2001. House prices were relatively flat from 1998 to 2001, suggesting that there was not a shortage of dwellings.

Table 5: Analysis of shortage or surplus of dwellings

June Years	Change in Population	Number of houses required if household size is 2.7	Number of houses required if household size is 2.5	80% of building consents	Surplus if size 2.7	Surplus if size 2.5
1995	53,300	19741	21320	18588	-1153	-2732
1996	58,600	21704	23440	17754	-3950	-5686
1997	49,400	18296	19760	18078	-218	-1682
1998	33,600	12444	13440	19390	6946	5950
1999	20,100	7444	8040	18086	10642	10046
2000	22,700	8407	9080	19395	10988	10315
2001	22,700	8407	9080	15476	7069	6396
2002	58,600	21704	23440	18026	-3677	-5414
2003	70,100	25963	28040	23259	-2704	-4781
2004	52,200	19333	20880	26601	7267	5721
2005	37,500	13889	15000	21955	8066	6955
2006	40,600	15037	16240	20450	5413	4210
2007	43,158	15984	17263	21230	5246	3967

 

Sources: Statistics New Zealand

Table 5 shows a shortage of dwellings at the aggregate level in 2002 and 2003. This shortage appears to have been more than filled by surpluses in the following years, suggesting that there was no shortage of dwellings by June 2007, regardless of whether or not an average household size of 2.5 or 2.7 is used in the analysis.

Household size is likely to be, at least in part, influenced by housing supply, affordability and house price increases. When prices are increasing sharply, the rate of household formation may slow; for example, young people may live with their parents for longer. The increase in average household size in the 2006 Census, after a long period of declining average household size, may suggest that such an impact has occurred. The analysis in table 5 is sensitive to the assumption about household size – it may be that the desired household size is smaller than the actual size. The numbers in table 5 can be recalculated with any desired household size. If the desired household size is assumed to be 2.3 people per household then there appears to be a small shortage. A household size of 2.3 is lower than the Statistics New Zealand projection of the expected average household size by 2021.

The numbers are also sensitive to the assumption that 80% of building consents end up as new dwellings. For example, if an assumption of 70% is used instead, there appears to be a small shortage. Between the 2001 Census and the 2006 Census the total number of dwellings increased by around 125,000, while there were just under 138,000 new consents issued over the same period. This calculation suggests 90% of consents were converted into dwellings, indicating that the 80% assumption used in table 5 is a conservative estimate.[12] As consents tend to lag behind actual demand it may be more appropriate to consider the consents with a six month lag. After making this adjustment the calculation suggests just under 90% of consents were converted into dwellings.

These estimates provide an approximation of the extent of undersupply or oversupply of housing. Under plausible assumptions, approach 1, suggests that there is no sign of a shortage of dwellings at the aggregate level. There are some limitations regarding the assumptions used, and the analysis does not pick up any regional dimension to the growth in dwellings and the population. For example, it may be that the increase in dwellings has come in places with little population growth, perhaps through growth in holiday homes. This appears to have been particularly important in coastal locations and the so-called sunbelt. The analysis also does not show how much of the construction has been of occupied dwellings versus unoccupied dwellings, which again may be holiday homes. These factors would all mean that a greater number of dwellings would be required to house the population, increasing the possibility of a shortage emerging. The use of a conservative estimate of 80% of building consents being converted into dwellings builds in an allowance for the construction of holiday homes that are not occupied by households on a regular basis.

Approach 2: Regional growth in occupied dwellings and population

The graph below provides an alternative approach to considering whether the increase in supply has been sufficient. It shows the population and occupied dwelling growth rates by region in the periods from 1996 to 2001 and 2001 to 2006. Unoccupied dwellings are initially excluded from the analysis as these dwellings are not available to house the population in a region, either because they are holiday homes, being renovated or left vacant for some other reason. The diagonal line shows the growth in occupied dwellings that would exactly match the population growth, implying no change in average household size. As a result, points above the diagonal line represent areas where dwelling growth has not kept up with population growth. In the 1996–2001 period, there were no regions where population growth exceeded dwelling growth. In the 2001–2006 period, there were 10 regions where dwelling growth was less than population growth, with the largest shortages in Auckland City, Manukau and North Shore City.

Figure 16: Growth in population and occupied dwellings 1996-2001 and 2001-2006

Source: Statistics New Zealand

To maintain average household sizes at their 2001 level of 2.6 people per household, the Auckland region would have needed around 4,500 extra dwellings, with Manukau accounting for half of these. The total shortfall across the rest of New Zealand is around 1,500 dwellings. An alternative method of calculation is to carry out this analysis using household growth rather than population growth. Using population growth, however, provides more flexibility to consider different desired average household sizes.

Reductions in average household size would shift the 45-degree line towards the x-axis. For example, if household formation has been suppressed by high house prices, the line would shift down, meaning that a larger number of regions would have a shortage of dwellings, and regions that already had a shortage would have a larger shortage.

While this analysis is at a regional level, it does not capture dynamics at finer levels of detail. For example, Auckland is a broad category, and there may be areas within Auckland where additional new dwellings are well below population growth – this could happen if a lot of the new construction was occurring in relatively expensive areas.

Figure 17 carries out the same analysis using total dwellings instead of unoccupied dwellings. The results are broadly the same, with most regions not showing any shortages, and the most substantial shortages emerging in Manukau and the North Shore.

Figure 17: Growth in population and total dwellings 1996-2001 and 2001-2006

Source: Statistics New Zealand

Type of supply

A further dimension of the increase in occupied dwellings is the type of dwellings that have been built. According to the 2006 Census, around 40,000 of the increase in occupied dwellings between 2001 and 2006 was non-separate buildings.[13] The proportion of non-separate buildings of the total stock of occupied dwellings increased from 16.9% in 2001 to 18.1% in 2006. Non-separate buildings are likely to have been an important dimension of the growth in supply in Auckland. Non-separate buildings, such as apartments, may not be appropriate for some households, particularly multi-family households. The growth in non-separate buildings may reflect increasing demand for other parts of the market, including small families, couples without children, single people and students, for smaller units with smaller sections to maintain. It may also reflect the magnitude of house price increases and a desire amongst households to find more affordable accommodation.

There is also evidence of a general upward trend in the size of dwellings that have been constructed. Building consents data show that the average floor size of new construction has increased from around 170 square metres in 2001 to 190 square metres in 2007. This increase has come despite the growing share of apartments, which will tend to be smaller in size than separate houses. The average floor area for multi-unit dwellings, or separate buildings, has been approximately stable at around 100 square metres, with the size of separate dwellings increasing more quickly.

Bigger houses tend to cost more than small houses; therefore, the trend towards new supply focussing on larger houses is likely to reflect demand at the upper end of the market. This may reflect larger returns to builders from constructing larger houses, as well as declining affordability for people on lower incomes. It may also reflect an existing availability of smaller houses such that purchasers of such houses do not need to build a new one. It does, however, mean that the supply of new dwellings is not meeting the needs of low-to-middle income households.

Figure 18: Type of occupied dwelling	Figure 19: Building consents average floor size
Source: Statistics New Zealand	Source: Statistics New Zealand

Conclusion: magnitude of supply increase

The two approaches used here have a number of drawbacks and provide mixed evidence regarding whether there is a shortfall in the growth of the number of occupied dwellings required to house the growing population during the boom in house prices. A small rise in average household size suggests that some household formation may have been delayed by house price increases. These two approaches, as well as a simple comparison of the change in the number of dwellings and the change in the population, provide little evidence of a widespread shortfall in the number of dwellings at the aggregate level. However, there are signs that shortfalls have emerged in the Auckland region. These shortfalls are likely to have added to price pressure and led to some increase in average household size.

As further indicators, relatively flat rent price changes in recent years and little change in the housing crowding statistic reported in the 2007 Social Report (Ministry of Social Development 2007) also support the hypothesis that the increase in the dwelling stock is close to that required to house the population. Again, there are some regions where the story is not so straightforward, with some levelling off of the long-term decline in crowding in Manukau. This result is consistent with the lift in household size and the apparent shortfall in the supply of new dwellings in Auckland.

While the trend in the growth of dwellings broadly matches, and in some cases exceeds, the growth of the population, the type of supply being generated does not meet the needs of all segments of the market. Much of the supply of new occupied dwellings has been in either non-separate buildings, including apartments, or large new houses. This has not met demand for smaller separate dwellings among low and middle-income groups. The general lift in prices has also meant that the existing dwellings that are available for these groups to buy have become more expensive and less affordable. As a result, the new supply has tended to be purchased by investors and then rented out. Even in areas where the supply of new dwellings has exceeded population growth, prices have risen sharply, reducing the supply of dwellings available for people on lower incomes.

7.2 Costs of building a house

The construction sector has responded to higher demand, however, this extra supply has come at a cost, with the costs of sections, materials and labour all increasing sharply during the boom in house prices. Section 7.2 discusses a variety of measures of the costs of sections, materials and labour.

Indices of construction costs

Statistics New Zealand data provides the most consistent measure of the increase in the cost of new building and construction. Statistics New Zealand provides a variety of measures of the costs of construction from the CPI and the Capital Goods Price Index (CGPI). There are a number of different sources of data regarding the changes in the cost of building and construction. These indices are compiled following internationally developed standards to ensure that the measure records a constant quality and type of construction over time. Between 2001 and 2007 these indices point to construction price increases of around 45%. The costs of construction have increased at a substantially faster rate than the total CPI. The trends in these series are shown in figures 20 and 21. These do not include the cost of land, but do include margins received by builders and developers as they measure the price paid by consumers.

The CPI aims to exclude any changes in price arising from changes in quality. A price index is created from a survey of builders that construct standard-plan houses. Respondents are asked to provide a quote for a house plan that they build fairly regularly. The survey is designed to measure pure price change by asking for the price of a specified plan on a level section, with land costs excluded from the calculation. To ensure that pure price change is measured, adjustments are made for changes in quality such as the size (floor area) of the plan and quality difference in fittings and materials. Changes to the building consents regime have in some cases resulted in improvements to the materials used in constructing, and the quality of, new house plans being tracked in the CPI survey. The impact these improvements had on surveyed prices was stripped out of the CPI index for purchase and construction of new dwellings (Statistics New Zealand, 2006).

Figure 20: Construction costs	Figure 21: Change in construction costs 2001-2007 compared with total CPI
Source: Statistics New Zealand	Source: Statistics New Zealand

Department of Building and Housing cost estimates

Direct estimates of building costs are also available. The Department of Building and Housing provides a six-monthly overview of the estimated costs of building a house.[14] The Department provides building cost data to help territorial authorities, and others, arrive at realistic estimated values of building costs – particularly for local authorities when they have questioned the job value provided with a consent application. Maltby and Partners Ltd, a firm of construction cost consultants, provides the Department with costing information for a series of building types. Maltby quotes from a set of authentic construction documents in order to establish a unit cost that is as accurate as possible. It should be noted that the Maltby data includes margins received by the building industry. The data does not fully account for changes in quality so as a result using better materials or fittings will show up as an increase in price.

The costs provided are for one-off houses, chosen in line with popular building practice in New Zealand. These costs do not reflect the economies that may be gained by builders building large numbers of similar houses nor do they reflect the additional costs normally associated with architecturally designed houses. To differentiate, group houses have been assessed as being, on average, 21% cheaper than one-off houses, while architecturally designed houses are assessed as being around 20% more expensive. The estimated costs include all necessary internal and external finishes to achieve compliance with the Building Code, all services and provision of standard appliances, and site works. Estimates are provided for two types of house, a house of 145 square metres and a house of 202 square metres house. These costs are summarised in table 6. The table also includes an additional row for median section prices to provide an estimate of the total cost of building a house.[15]

Table 6: Costs of building a house

	2001	2007	$ increase	%increase
Building cost
145 mtr2	150,607	247,636	97,029	64.4
202 mtr2	192,708	292,631	99,923	51.9
Land Cost
Median section	81,250	175,000	93,750	115.4
Total Cost
145 mtr2	231,857	422,636	190,779	82.3
202 mtr2	273,958	467,631	193,673	70.7
Building share of total cost (%)
145 mtr2	65.0	58.6
202 mtr2	70.3	62.6
Section share of total cost (%)
145 mtr2	35.0	41.4
202 mtr2	29.7	37.4

 

Source: Department of Building and Housing, Real Estate Institute of New Zealand

The estimated increase in building costs of 64.4% for the house of 145 square metres and 51.9% for the house of 202 square metres are both higher than the increases recorded in the Statistics New Zealand indices discussed above. Some of the reason for the difference may be accounted for by changes in quality, with the DBH figures not fully accounting for quality changes. The smaller house of 145 square metres has a higher concentration of service areas, such as kitchens and bathrooms, than the house of 202 square metres.

The smaller of the two houses is likely to be more relevant for first home buyers because of the lower cost. The Unit’s focus on affordability means that trends in the cost of the smaller type of house was the primary interest.

As part of the project, Maltbys provided the Unit with a more disaggregated breakdown of the costs of the smaller of the two modal houses. Table 7 summarises this more detailed breakdown.

Table 7: Costs of constructing a house of 145 square metres

	2001	2007	$ increase	% increase	Share of total cost 2001	Share of total cost 2007
Median Section Price	81250	175000	93750	115.4	34.7	40.8
Cost of materials	94271	143562	49291	52.3	40.3	33.5
Cost of labour	58517	110260	51743	88.4	25.0	25.7
Total	234039	428823	194784	83.2	100	100

Source: Department of Building and Housing, Real Estate Institute of New Zealand

New dwellings are a relatively small share of total house sales, accounting for around 20% of annual sales. Nevertheless, households wishing to upgrade their homes and first home buyers can substitute between new houses and existing houses. Trends in the costs of new houses can influence the prices of existing houses and vice versa.

Figure 22: Components of costs of 145 square metre house	Figure 23: Costs and house prices
Sources: REINZ, Department of Building and Housing	Sources: REINZ, Department of Building and Housing

7.3 The cost of land has doubled since 2001

New Zealand cities have historically grown by expanding the land area that they cover, with new suburbs housing a growing population. As a result, New Zealand cities have population densities that are below those in cities in much of the world, although closer to densities in most Australian cities. Local governments, particularly in Auckland, have sought to control growth in the land area covered by cities, largely in order to limit the infrastructural and environmental costs of expansion. In Auckland the expansion of the city is governed by the Metropolitan Urban Limit (MUL). The House Prices Unit was unable to ascertain the extent to which the MUL may have influenced land prices in Auckland – many of the factors that have driven land prices upward would exist irrespective of the MUL and the Unit was not able to adequately isolate which price effects may have resulted from the application of the MUL compared with other factors.

Raw land
An important dimension of land supply is the extent to which there is an adequate supply of land ready, or close to ready, for construction. There is a supply pipeline for housing that begins with land that is in other uses and available under district plans for housing use and land already in housing use that is available under district plans for more intensive housing use. That pipeline extends into a further stage when housing consents are issued and the land is developed for housing. A stock of 15–20 years worth of land supply does not provide any information about how much of that land is ready for construction. If very little of it is ready for construction, the price of land that is ready for construction is likely to appreciate.

The practical requirements for contiguous development mean that in situations of high demand, land adjoining cities is likely to appreciate in price because it is likely to be easier and cheaper to develop than converting more geographically distant land and supplying the necessary infrastructure. The quantities of land available ready for development and ready for construction are therefore important determinants of prices.

New Zealand has no data sources about the state of land, the quantities at various points of the development process and the prices of that land. As a result, it is difficult to reach strong conclusions about land supply issues. Increasing the availability of such information is a crucial component of improving the quality of monitoring and regulating land supply. Obtaining more information about land and its current status should be a priority for future work.

The Australian Productivity Commission (2004) notes that the establishment of a boundary reduces the risk of investing in designated areas, and signals that services are likely to be made available in the future, both of which will enhance the value of the land. The Productivity Commission concluded that constraints on the supply of land at the urban fringe had contributed to housing price pressures, particularly in Sydney, by increasing the scarcity value of land. It also concluded that, “...because recent price increases have been due mainly to the surge in demand in established areas, improvements to land release policies or planning approval processes could not have greatly alleviated them.”

Grimes and Liang (2007) conducted a detailed econometric analysis of the impact of urban boundaries in Auckland. They concluded that land just inside the boundary from 2001 to 2003 was between 4.5 and 13 times more expensive than land just outside the boundary, with the size of the impact varying depending on the estimation technique used, and that Auckland prices had risen substantially relative to other regions in the North Island. All but one of the measures indicated a ratio of at least 10 in the final years, leading the authors to conclude that the ratio was towards the high end of the 4.5 to 13 range. Grimes and Liang also found that land just outside the boundary had recorded the largest increases between 1993 and 2003, perhaps due to an expectation of future relaxation of the growth limit.

Other cost drivers

Other factors beside the raw land cost also influence final section prices – the costs of finance, development levies and the costs of development are a larger component of section prices than the raw land itself.

Table 8: Factors increasing the price of a developed 400mtr² section on the edge of Auckland

Factor	Impact on the change in developed section price between 2001 and 2007
Increases in the price of land sold in blocks larger than 1 hectare	Estimated very indicatively to have increased by 300% noting further analysis would be needed to verify this estimate.
Council-imposed developer levies	Estimated very indicatively at an additional $25,000, but varies with area and can be as much as $40,000 in Auckland city.
Higher cost of capital (including holding costs from regulatory delays)	Estimated to have increased by around 50%.

7.4 Costs of materials have increased

The Maltbys data suggests that the costs of almost all materials have increased substantially. The largest increases are for electrical, carpentry and plumbing materials, which together account for around 50% of the total increase in the cost of materials.

The rising prices of materials partly reflect the international price of these inputs as well as the margins of local distributors. There are only two large-scale providers of building materials in New Zealand, and during a construction boom these providers may be able to increase their margins.

7.5 Labour costs have increased

Trends in labour costs
There are a variety of measures of labour costs that are available. The Labour Cost Index (LCI) controls for the quality and quantity of labour used shows that nominal hourly wages increased 28% in the construction industry and 14% in the economy as a whole between March 2001 and June 2007. The LCI only covers wage and salary earners, so it does not capture changes in the earnings of self-employed people, which is an important source of labour in the construction industry.

Linked Employer-Employee Data (LEED) compiled by Statistics New Zealand show that median nominal earnings of self-employed people in the construction industry increased 39% between 2000 and 2006. The nominal median earnings of wage and salary earners in the construction sector have increased 20% over the same period. The increase in nominal earnings in the construction industries among wage and salary earners is close to the 19% increase across all industries over the period, however, among self employed people, the 39% increase in construction is well above the 27% increase in all industries.

Table 10: Median earnings

 

Construction				Total Industry
Main income source	Wages and Salaries	% change	Self-Employment	% change	Wages and Salaries	% change	Self Employment	% change
Tax Year
2000	28,970		25,470		25,680		24,330
2001	30,030	3.7	25,260	-0.8	26,190	2.0	26,410	8.5
2002	31,320	4.3	27,570	9.1	27,030	3.2	28,600	8.3
2003	31,770	1.4	30,330	10.0	27,690	2.4	28,030	-2.0
2004	32,670	2.8	32,500	7.2	28,650	3.5	29,230	4.3
2005	33,340	2.1	34,380	5.8	29,510	3.0	30,290	3.6
2006	34,930	4.8	35,320	2.7	30,570	3.6	30,840	1.8

 

Source: Statistics New Zealand (LEED data)

The Maltbys data also suggests that the cost of labour involved in building a house has increased. The cost increases reported in the Maltbys’ data are well above the increases recorded in the official Statistics New Zealand measures of the changes in earnings of people working in the construction industry. One possible explanation for this observation is that the Maltbys data includes the margins received by building firms and not just the wages received by people working in the industry. This suggests that margins have increased during the boom. Despite some conflicting information from the different data sources, both sources show that labour costs have increased by more in the construction sector than in the rest of the economy.

Training, skill shortages and labour demand

Rising labour costs reflect, more generally, the construction sector’s increasing demand for labour during a boom in construction activity and the availability of skilled and unskilled labour to work in the sector. The Department of Labour carries out skill shortage assessments as part of its Job Vacancy Monitoring programme. In June 2006, the Department examined 14 different trade occupations and assessed all 14 trades as having genuine skill shortages, where employers have difficulties filling their job vacancies because there are not enough people with the required skills in the labour market to fill the positions on offer. Amongst carpenters, only 34% of positions were filled within 10 weeks of advertising, and there was an average of only eight suitable applicants for every ten carpenter vacancies.

The residential construction sector competes with other parts of the construction sector, which have also been expanding quickly in recent years, with strong growth in commercial and infrastructure projects, such as roads, prisons and hospitals. Non-residential construction work put in place increased 23.0% in the year to June 2007 compared with the year to June 2007. Residential work put in place increased 40.0% over the same period. Table 11 shows the strong growth in employment in the construction industry, based on LEED data, from 2000 to 2006.

 

Table 11: Employment growth

Construction			Total Industry
	Total number of people	% change	Total number of people	% change
Tax Year
2000	122,349		1,859,400
2001	122,478	0.1	1,902,660	2.3
2002	124,476	1.6	1,947,339	2.3
2003	133,158	7.0	2,014,878	3.5
2004	145,254	9.1	2,086,851	3.6
2005	160,845	10.7	2,166,081	3.8
2006	173,196	7.7	2,230,170	3.0

 

Source: Statistics New Zealand (LEED data)

A decline in the training of apprentices in the trades during the 1990s diminished the available pool of workers for the construction industry. For example, the number of fitters and turners between the ages of 20 and 24 declined from nearly 1,150 in 1991 to around 200 in 2001. More recently, there has been an increase in training for qualifications related to the trade occupations, with the number of trainees enrolled doubling between 2001 and 2005. The carpenter occupation was the largest contributor to this increase. All occupations recorded an increase over the 2001 to 2005 period, and most recorded growth of over 10%.

Effect of migration

Immigration policy is also a significant determinant of the availability of skilled workers in the construction industry. Policy changes in 1991 and 1995 altered the way people gained points for residence. A focus on qualifications made it harder for those from the construction sector to be approved for residence, with arrivals falling to 900 in late 1998 after peaking at close to 1,500 in 1996.[16] The recent introduction of the Skilled Migrant Category, where the focus is on skills and employment, has contributed to a lift in arrivals of construction workers in the past two to three years, to around 1,800 per year. The annual number of departures of construction workers from New Zealand, which has risen to 2,000 in the past 12 months, compared with around 1,400 in 2002, is also a key determinant of the availability of skilled labour.

Figure 24: Net migration –total and construction workers	Figure 25: Construction share of arrivals and departures
Source: Statistics New Zealand	Source: Statistics New Zealand

7.6 Regulation

There are three major regulatory areas that affect housing:

• The Resource Management Act provides for the sustainable management of land.
• The Building Act requires building consent approval before building can commence.
• Development levies to be paid under the Local Government Act.

Resource Management Act

There are two major features of the compliance system faced by developers under the Resource Management Act (RMA):

• There are costs associated with the resource consent notification and consultation processes for full discretionary or non-complying activities. In these cases, the developer is required to notify and consult with the public, and, may have to take a case to the Environment Court.
• Time delays in gaining a resource consent regarding the use of land and increased asset holding costs for developers may contribute significantly to the final cost of building a house.

Analysis by the Ministry for the Environment of resource consents in 2005/06 showed that:

• A total of 99.3% of all consent applications are approved
• A total of 94% of all consent applications are non-notified
• Of 52,000 consents processed in 2005/06, only 357 were declined
• Only 1% of all consents are appealed to the Environment Court.

The small number of notified consents is consistent with one of the conclusions of Grimes et al (2006b), where developers reported that notifiable consents were most at risk of delays from objections and subsequent hearings. As a result, developers reported that they would avoid developments that required consents that needed to be notified, thereby avoiding some potentially innovative projects that might make the development notifiable.

The Ministry for the Environment (2007) has analysed the processing of consents under the Resource Management Act. Local authorities were asked for the number of resource consents of each type processed within statutory time limits. This includes resource consents where the time limits were formally extended by local authorities under section 37 of the RMA.

In 2005/06, 73% of all resource consents were processed within statutory time limits. This compares with 77% in 2003/04 and 82% in 2001/02. Compared to the previous survey, water and discharge consents increased in the proportion processed within time. The other categories (subdivision, land-use and coastal) all decreased in the proportion processed within time. It is important to note that in these processing results statutory timeframes are calculated as the number of working days between the date a local authority receives a resource consent application and the date it releases its decision on that application. The time it takes for an applicant to supply further information in response to a request from a local authority is not counted, nor are appeals or objections to decisions made on consents. As a result, in some cases the actual elapsed time may be longer. Data on the actual elapsed time, which is the important factor for a developer who incurs holding costs, is not available.

Local authorities were asked for the number of notified, limited notification or non-notified consents that they processed within statutory time limits.[17] This includes resource consents where the time limits were formally extended by local authorities under section 37 of the RMA. In 2005/06, 56% of notified resource consents were processed within statutory time limits. For limited notification consents it was 60%, and for non-notified consents it was 74%. Regional councils processed 86% of resource consents within statutory time limits compared with 71% for territorial authorities and 58% for unitary authorities.

The Ministry for the Environment has identified a number of reasons for delays, including:

• inadequate information regarding environmental effects in submitted application
• a high number of complex applications
• a lack of local authority processing capability due to shortages of skilled staff
• difficulty obtaining expert and legal advice in small local authorities
• increased complexity, requiring additional technical and engineering input, as development takes place on land previously considered uneconomic.

The Unit has not had the time or resources to carry out a full review of the issues associated with the RMA. Results from a pilot study completed in by the Ministry for the Environment in November 2007 indicate that consent applicants had difficulty in differentiating between costs charged by the council for processing the consent, the costs of getting building consents for the same development, development contributions under the Local Government Act and regular design costs that would have been incurred regardless. Of the 24 consent applicants interviewed for the pilot, eight cited holding costs as an issue.

The Ministry for the Environment proposes to undertake research in 2008 to objectively measure the costs to applicants of obtaining a resource consent. That research will provide a better information base on the effect of land use regulations on housing costs.

Building Act

The Building Act 1991 introduced a performance-based approach to building regulation. It established the Building Industry Authority to develop the Building Code and provided Territorial Local Authorities (TLAs) with the responsibility for approving alternative solutions, issuing building consents and undertaking inspections during and at the completion of construction. Failure of the building system, including design, construction and inspections, led to the leaky buildings failures, which resulted in an amendment to the Act in 2004.

A new regulatory framework was introduced with the Building Act 2004. This lifted the regulatory burden on the construction sector, in part to address issues arising from the failure of the system to prevent leaky buildings. An estimate of the economic costs carried out as part of the assessment of the regulatory impact of the Act suggested that it would add 2.9% to the cost of building a new dwelling.[18]

Developers report that building consent approval times have increased in recent years.[19] Factors affecting the delays include:

• innovative development proposals being treated more cautiously by authorities, due to the leaky homes problems
• inadequate staffing volumes, lack of experienced staff and recruiting difficulties in a tight labour market.

Comprehensive data about the time taken by councils to approve building consents is not currently available, however, the Department of Building and Housing is working with Local Government New Zealand to assemble time and cost data from individual TLAs. Results from 36 of the 73 councils surveyed show that of the building consents logged and closed in the 2006/07 fiscal year, 85% met the 20 working days stipulated in the Building Act 2004. This rate varies among the 36 councils analysed. Chatham Islands Council has 100% of its 15 building consents meeting the 20 working days requirement. On the other hand, Manukau City Council processed 53% percent of its building consents within the 20 working days requirement. The variation in consent processing time can be attributed to factors such as the nature and value of building consents and the number of consents councils are processing in relation to staff numbers and other supporting resources available.

The data above regarding the time taken to process consents does not measure the total time elapsed in processing consents. The statutory time excludes time where an application is suspended because of a lack of information and returned to the applicant. The analysis also only includes those consents logged and closed in the 2006/07 year so does not include any long standing applications that were still open at the end of the year. These limitations make it difficult to use this data to get a clear picture of the impacts of the time taken to get a building consent.

Among those building consents that took longer than 20 working days to process, 45% of them were processed in the next five working days. There were still a small number that took longer to complete, and these tended to be for the large building projects.

Central government has imposed a more rigorous regulatory regime through occupational licensing and building consent authority accreditation, however these costs are relatively small in the context of the price of building a house. The requirement in the Building Act 2004 for Building Consent Authority accreditation is intended to help councils improve the quality of their systems and processes which, over time, are intended to reduce delays in building consent processing and improve consistency. The costs of the accreditation process have not been evaluated.

Council imposed development levies

Council infrastructure levies are a sizable component of new housing construction costs and they appear to be rising. Purchasers of new houses expect a range of services to come with the house, including sewerage, drainage, water and electricity. Councils are increasingly charging developers for the provision of these services rather than funding them out of local authority rates or the central government funding from general taxation.

There are two broad areas for which charges are sometimes levied. These are:

• Economic infrastructure: services such as water and sewerage.
• Social infrastructure: a range of community and recreational services, such as parks and libraries.

The Unit has not attempted to consider issues regarding who should pay for infrastructure or what the quality of infrastructure should be. There is considerable variability between territorial authorities as to where the charges are applied and the size of the charges. Further work is needed to understand the impact that the variability and levels of infrastructure contributions between territorial authorities have on the cost of new housing.

Summary of regulatory costs

The regulatory costs identified by the House Prices Unit are estimated as:

• Direct costs to developers arising from development levies of $5,000–$40,000 per dwelling, depending on location.
• Building consent fees of $1,500–$3,000 per dwelling, Building Consent Accreditation costs of around $10 per dwelling, as well as a one-off accreditation fee of $15–$50,000, depending on region. In total, building consent costs are estimated to add 1–3% to the cost of a new dwelling.
• Occupational licensing of $400–$1,000 per licence, estimated at $100 per dwelling.
• Holding costs of around $15,000 per dwelling. These primarily arise from delays associated with planning approvals regarding land use under the Resource Management Act, but they can arise in the case of large building projects that take time to receive building consent approval.

While analysis by the Ministry for the Environment and the Department of Building and Housing suggests that most resource and building consents are processed in line with the specified timeframes, there are some, particularly large and/or complex applications for resource and/or building consents, that do face delays.

All of these costs add to the price of new dwellings. New dwellings account for around 20% of the total turnover of dwellings in New Zealand, so they are a relatively small part of the market. Nevertheless, a purchaser can substitute between buying a new dwelling or a second-hand dwelling, so changes in the costs of new dwellings can influence the wider market.

7.7 Design and building techniques

New Zealand has a highly customised and labour intensive approach to residential construction. Builders tend to be small businesses dealing with individual clients and their preferences, with most construction individualised or a relatively small variation on a core design. As a result, it is difficult to take advantage of economies of scale, and most work is carried out on site, rather than by assembling prefabricated materials that have been manufactured off site.

In the long term, moderation in house price growth would be assisted by sustainable cost reductions in the price of land, materials or labour. Sustainable construction cost reductions are likely to result from:

• Improved productivity and skills in the building sector. Productivity issues are addressed in more detail in box 5.
• Faster processing of building consents and approvals under the Resource Management Act.
• Increased intensification and returns to scale as the size of projects lifts.