• Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Potential ways to improve New Zealand's housing quality statistics

This chapter discusses potential ways of improving data collection of housing quality information in New Zealand. See:

Criteria for choosing options

Any expansion of housing quality statistics will require some investment. Any solution must focus on the most cost-effective and useful ways to improve the collection of housing quality data.

The criteria for choosing options includes likely quality of information, but also how feasible it is to collect the data.

An understanding of the user requirements for housing quality data is also vital before any programme for collecting housing quality is established, as this will shape the way the programme develops:

  • Is the purpose of collecting housing quality information merely to inform?
  • Or is the purpose to provide a basis for quality improvement within New Zealand’s housing stock? For example, housing quality information overseas (such as in Scotland) is used to improve the quality of housing, and the survey has tracked a gradual improvement in Scotland’s housing stock.

Options to consider

Here are eight options New Zealand could adopt to expand information on housing quality. We established these by looking at information collected nationally and internationally.

We ranked these options by the likely quality and usefulness of the information gathered, where five stars are highest and one star is lowest. Purpose-built survey based on an overseas model such as the Scottish or English housing condition surveys. See:

Purpose-built survey based on an overseas model such as the Scottish or English housing condition surveys 

Rating: * * * * *

Type of data source: Survey

Pros

  • Would be the most flexible option as it could be designed to fit the purposes of collection.
  • Would provide high-quality information.
  • Could use some of the existing infrastructure, such as BRANZ.

Cons

  • Likely to be the most expensive option.
  • English and Scottish survey models are based on legislation and aim to improve the quality of housing stock. We would require a similar legislative mandate in New Zealand.

Rationale for this option
A purpose-built survey would allow the most flexibility and scope for collecting housing quality information. The potential for including information on a wider range of housing, including uninhabitable housing, is high. This option is likely to be more expensive, but would provide the highest-quality information.

Expanded BRANZ survey linked to a household survey

Rating: * * * * *

Type: Survey

Pros

  • Would utilise existing surveys with established methodologies.
  • Would link physical/objective data with demographic characteristics and self-reported housing quality.
  • Would be able to compare housing quality for different ethnic groups and household types.
  • Evidence shows that homeowners/renters overestimate housing quality and therefore self-reported housing quality data is not as reliable.
  • BRANZ survey has time series so it can compare housing quality over time.

Cons

  • Important to include renters in sample but BRANZ found this difficult in their previous survey.
  • Would require funding – housing inspections are expensive.

Rationale for this option
This option links physical/objective data with demographic characteristics and self-reported housing condition. This would be a preferred option as it would generate high-quality information.

Every five years BRANZ carries out a housing condition survey. Before 2010, this survey focused on Auckland, Wellington, and Christchurch and was confined to owner occupiers. The Centre for Research, Evaluation and Social Assessment and the Department of Building and Housing funded a nationwide survey in 2010, which also included a sample of rental housing. The sample for the entire survey was fairly small, totalling 491 houses. Given the disparity between self-reported housing quality and objective assessments, linking the BRANZ data with a household survey would be very valuable.

Expanded BRANZ survey with larger sample size and greater range of demographic and socio-economic questions

Rating: * * * *

Type: Survey

Pros

  • Would provide better subnational coverage.
  • Existing methodologies.

Cons

  • Unlikely to be a large enough sample to compete with the depth of information from a household survey.
  • Could be a costly option depending on to what extent the sample was extended.

Rationale for this option
An expanded BRANZ survey with a larger sample size and a greater range of demographic and socio-economic questions is another potential source of housing quality data.

EECA: including a housing inspection/questionnaire with applications for the subsidised insulation scheme (Warm Up New Zealand: Healthy Homes)

Rating: * * *

Type: Survey/administrative

Pros

  • Would provide more information about housing for low-income people.
  • Link with existing insulation programme would be cost efficient.
  • Questionnaire would be cheaper but quality might not meet user needs.

Cons

  • Would not be a representative sample of dwellings.
  • Building inspection would provide the best results, but could be expensive unless the inspector uses a minimum checklist.
  • Would require consistency of application in order to get reliable results.

Rationale for this option
Warm Up New Zealand: Healthy Homes is a three-year government insulation programme focused on low-income households, beginning in 2013. It will deliver about 46,000 warmer, drier, and healthier homes. The programme will be targeted at households (including renters) that have a Community Services Card and are at high health risk (EECA, nd, a).

Rental warrant of fitness

Rating: * * *

Type: Administrative data.

Pros

  • This scheme is currently being developed by Housing New Zealand in conjunction with other agencies.
  • Would be cost effective to use existing data collection.
  • Will be applied to Housing New Zealand properties first, but has the potential to be expanded to all rental properties.
  • Rental properties are acknowledged to be of poorer quality, so this scheme would provide valuable insights into this market segment.
  • Would be most useful if it could be linked to a household survey / census, in order to look at characteristics of inhabitants.

Cons

  • As the scheme has not yet been established there are a number of unknowns.
  • It will not cover owner-occupied properties.

Rationale for this option
One proposal to improve the quality of rental housing in New Zealand is to introduce a compulsory warrant of fitness for housing. The Government is currently looking at the feasibility of imposing a warrant of fitness on rental housing, as proposed in December 2013 by an expert group appointed by Children’s Commissioner Dr Russell Wills. If this proposal was to go ahead it should include an element of data collection for statistical purposes to:

  • enable assessment of effectiveness of the scheme for improving housing quality
  • improve the collection of housing quality data.

Expand housing information asked in a household survey

Most likely the New Zealand General Social Survey (NZGSS) or Household Economic Survey (HES).The Household Labour Force Survey is an unlikely option, but would provide the best subnational coverage.)

Rating: * * *

Type: Survey

Pros

  • Would use existing surveys.
  • NZGSS already includes a housing quality section.
  • HES has information on housing costs and maintenance.
  • A cheaper option because it would use self-reported housing quality.
  • Would generate national and limited subnational statistics, and enable disaggregation by different groups.

Cons

  • Self-reported housing quality has been shown to be less reliable and significantly under-reports housing issues.
  • Limited subnational disaggregation, but potential for data pooling to be used to generate better subnational statistics.

Rationale for this option
Expanding the housing information in household surveys would be the simplest option as the vehicles for collection already exist and two household surveys already collect some housing information. This option would allow the topics identified by the OECD and the 2009 Review of Housing Statistics to be covered. Three household surveys have potential for data collection on housing quality. However, since the Household Labour Force Survey is an unlikely option, we have not discussed it here.

New Zealand General Social Survey (NZGSS)
Statistics NZ carries out the NZGSS every two years, and already collects some information on housing quality. Through the NZGSS, housing quality information can be evaluated in relation to mental and physical health, as well as housing satisfaction. It would also be useful to investigate the option of harmonising some questions with Australia in order to compare self-reported housing quality in the two countries.

The NZGSS sample size in 2010 was 8,550 people. The NZGSS can be broken down by region as well as by age, ethnicity, and deprivation. It would be useful, however, to also include some housing affordability information.

Household Economic Survey (HES)
Statistics NZ carries out this survey every three years, but in the two years between the three-yearly HES, a shortened version of the survey (HES (Income)) is conducted to collect information on income and housing expenditure. Since HES already collects information about housing, an added supplement on housing quality to HES (Income) would also prove feasible. However, HES is the longest and most complex survey to answer, so it would only be possible to include a limited range of questions. This survey has the added benefit of housing affordability information (housing costs and income). HES has a fairly small sample size of 4,700 households and can therefore only provide limited regional or ethnic breakdowns.

Household survey information with links to other data sources

Rating: * *

Type: Survey/administrative

Pros

  • Would utilise existing data sources, so it would be cheaper than physical inspection of properties, depending on the cost of QV data.

Cons

  • QV housing quality data is considered old and unreliable (housing quality is now no longer assessed). The BRANZ survey shows very poor linkage between QV ratings and physical inspection data.
  • We do not recommend this option, although there is potential to expand other types of housing information from QV data (such as material of dwellings, floor area).

Rationale for this option
Self-reported housing quality data would be more valuable if there was a way to link this information with other data sources. Currently, however, there is limited administrative information available. QV provides some information, but the housing quality information is very limited and considered out of date and unreliable.

Limitations of QV data on housing quality
QV New Zealand maintains a large dwelling information database, which includes some information about the physical characteristics of the dwelling, including age and materials used for walls and roof. Two fields in the database relate to housing quality: building condition and weather tightness risk, although these fields must be used with some caution. For example, in most cases these assessments of risk for weather-tightness issues are modelled based on information about materials and construction.

Buckett, Marston, Saville-Smith, Jowett, & Jones (2011) compared the inspections of BRANZ assessors with QV assessments (which are mostly exterior) and found a considerable difference in the assessment of quality. BRANZ assessors found that 44 percent of rental properties were in poor condition in 2010, compared with QV, which assessed only 5 percent of rental properties as poor. The most useful fields from QV are therefore likely to be the age of dwelling and materials used for walls and roof.

EECA Homestar rating system

Rating: *

Type: Self-selected sample

Pros

  • Existing information that is gathered at little cost.
  • Would require further investigation to find out feasibility of generating statistics.

Cons

  • Respondents are self-selected.
  • Self-reported housing quality under-reports issues.
  • Most New Zealand houses receive a rating of between two and three stars (out of a possible 10), so rating system is unlikely to include the granularity needed for comparison purposes.

Rationale for this option

This is a voluntary rating system questionnaire where householders answer a range of questions relating to the energy efficiency and sustainability of their home. After the questionnaire is completed the home receives a star rating and suggestions for improvement (see Homestar website). This data source could be investigated to see whether it could be used to generate statistics on housing quality.

Limitations
The data currently has a number of limitations. Currently, all respondents are self-selected and the database notes that most New Zealand houses receive a rating of between two and three stars (out of a possible 10). Therefore, it is unlikely to include the granularity needed for comparison purposes. 

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Top
  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+