Stats NZ has a new website.

For new releases go to

As we transition to our new site, you'll still find some Stats NZ information here on this archive site.

  • Share this page to Facebook
  • Share this page to Twitter
  • Share this page to Google+
Household Labour Force Survey: September 2014 quarter
Embargoed until 10:45am  –  05 November 2014
Data quality

Period-specific information
This section is for information that changes between periods.

General information
This section has information about data that does not change between releases.

Period-specific information

Achieved sample and response rate

In the September 2014 quarter, 31,164 people in 15,790 households responded to the Household Labour Force Survey (HLFS).

The target response rate for the HLFS is 90 percent. The response rate for the September 2014 quarter was 86.2 percent and the achieved sample rate was 76.0 percent.

General information

Data source

The target population for the HLFS is the civilian, usually resident, non-institutionalised population aged 15 years and over.

The statistics in this release do not cover:

  • long-term residents of homes for older people, hospitals, and psychiatric institutions 
  • inmates of penal institutions
  • members of the permanent armed forces
  • members of the non-New Zealand armed forces 
  • overseas diplomats
  • overseas visitors who expect to be a resident in New Zealand for less than 12 months
  • those aged under 15 years.

Accuracy of the data

Sample design

The HLFS sample contains about 15,000 private households and about 30,000 individuals each quarter. We sample households on a statistically representative basis from areas throughout New Zealand, and obtain information for each member of the household. The sample is stratified by geographic region, urban and rural areas, ethnic density, and socio-economic characteristics.

Households stay in the survey for two years. Each quarter, one-eighth of the households in the sample are rotated out and replaced by a new set of households. Therefore, up to seven-eighths of the same people are surveyed in adjacent quarters. This overlap improves the reliability of quarterly change estimates.

The period of surveying/interviewing is 13 weeks (eg from 6 July to 5 October 2014). The information obtained relates to the week before the interview (referred to as the ‘survey reference week’). We first interview respondents face-to-face at their home. Subsequent interviews are by telephone wherever possible. Respondents also have the option to file self-completed questionnaires.

Where practicable, we obtain information directly from each household member. Otherwise a proxy interview is conducted, in which we obtain details from another adult in the household.

Sampling errors

Sampling errors can be measured. They quantify the variability that occurs by chance because a sample rather than an entire population is surveyed.

We calculate sampling errors using the jackknife method. It is based on the variation between estimates of different subsamples taken from the whole sample. This is an attempt to see how estimates would vary if we were to repeat the survey with new samples of individuals.

We calculate sampling errors for each cell in the published tables and for estimates of change between adjacent quarters at the 95 percent confidence level. For example, if the estimated total number of people employed was 2,500,000, with a sampling error of plus or minus 50,000, or 2 percent, this means there would be a 95 percent chance the true number of employed people lies between 2,450,000 and 2,550,000.

Smaller estimates are subject to larger relative sampling errors than larger estimates. For example, the estimated number of Pacific peoples employed would have a larger relative sampling error than the estimated total number of people employed. Likewise, the estimated number of people unemployed would have a larger relative sampling error than the estimated number of people employed.

Estimates of change are also subject to sampling error. For example, if the survey estimate of change in total employment was up 50,000 over the year, and was subject to a sampling error of plus or minus 12,500 (25 percent), this means the true value of the change in surveyed employment would have a 95 percent chance of lying between 37,500 and 62,500.

A change in an estimate, either from one adjacent quarter to the next, or between quarters a year apart, is said to be statistically significant if it is larger than the associated sampling error. Therefore, the example quoted above represents a significant movement.

In general, the sampling errors associated with subnational estimates (eg breakdowns by regional council area or ethnic group) are larger than those associated with national estimates.

A non-sampling error is very difficult to measure, and if present can lead to biased estimates. We aim to minimise the effect of these errors by applying best survey practices and monitoring known indicators.

Response rate and achieved sample characteristics

The achieved sample size measure is the number of eligible households and individuals that responded to the HLFS in the quarter. The achieved sample size typically increases over time as the population grows and more dwellings are added to the survey sample.

We calculate the response rate by determining the number of eligible households that responded to the survey as a proportion of the estimated number of total eligible households in the sample.

The following table shows the HLFS achieved sample and response rates for the last five quarters.

 HLFS achieved sample and response rates
 Quarter National response rate (percent)  Achieved sample rate (percent)  Achieved sample Individuals  Achieved sample Households 
 Sep 2013 84.7 74.7 30,296 15,337
 Dec 2013 87.2 76.9 31,095 15,795
 Mar 2014 87.3 77.6 31,529 15,990
 Jun 2014 85.4 75.0 29,888 15,547
 Sep 2014 86.2 76.0 31,164 15,790

Obtaining a sample that represents the population is essential when it comes to producing reliable labour force estimates. The HLFS goes through three stages of weighting to achieve this.

See New quality measures for the Household Labour Force Survey for more information.

The following figure shows that while the distribution of the pre- and post-calibration weights differs within a quarter, the difference between the weights typically does not change from quarter to quarter.

Age distribution_qtr116

The undercoverage rate gives an indication of how representative the pre-calibrated sample is. The higher the undercoverage rate, the less representative the pre-calibrated sample.

Usually the undercoverage rate in the HLFS is around 20 percent. The overall undercoverage rate for the HLFS in the September 2014 quarter was 16.2 percent. This compares with 18.2 percent in the June 2014 quarter and 15.1 percent in the September 2013 quarter.

Using a proxy

Where practical, the HLFS gets information directly from each household member. Otherwise, we conduct a proxy interview in which details are given by another adult in the household.

The quality of data from proxy responses is affected by two factors: what type of information is being asked for, and the relationship between the proxy (the person that the survey questions are being answered for) and the proxy respondent (the person replying to the questionnaire on behalf of the proxy). More than 90 percent of related people answer correctly for key variables. When the proxy and proxy respondent are unrelated there is still a high quality of response.

We calculate the proxy rate as the percentage of respondents who had someone else respond on their behalf divided by the total number of respondents. A typical proxy rate in the HLFS is around 30–35 percent. This excludes quarters when a supplement was attached to the HLFS. When a supplement is attached the proxy rate typically falls. This is because supplements often have different proxy rules, which have a small effect on how HLFS responses are collected.

The proxy rate for the HLFS in the September 2014 quarter was 33.7 percent. This compares with 27.3 percent in the June 2014 quarter and 33.6 percent in the September 2013 quarter. Supplements are attached to the HLFS in June quarters.

Seasonal adjustment and trend series

In the labour market, cyclical events that affect labour supply and demand occur around the same time each year. For example, in the summertime a large pool of student labour is both available for, and actively seeking, work. Demand for labour in the retail sector and in many primary production industries also increases.

For any series, we can break the estimates down into three components: trend, seasonal, and irregular. Seasonally adjusted series have the seasonal component removed. Trend series have both the seasonal and irregular components removed, and reveal the underlying direction of movement in a series.

We adjust the series for each labour market statistic separately. For this reason, the sum of the seasonally adjusted estimates for employment, unemployment, and people not in the labour force will usually not add up to the working-age population estimates.

Seasonal adjustment has more information about how we seasonally adjust our statistics. Seasonal adjustment makes data for adjacent quarters more comparable by smoothing out the effect on the times series of any regular seasonal events. This ensures that the underlying movements in the time series are more visible. 

See 'Revisions' for information on the change in estimates between the current and previous publication for the seasonally adjusted and trend data.

All seasonally adjusted and trend series are produced using the X-13ARIMA-SEATS Version 1.1 package developed by the U.S. Census Bureau. The seasonal adjustment package for HLFS series was upgraded from X-12-ARIMA to X-13ARIMA-SEATS for the June 2014 quarter to comply with international best practice.

Prior adjustments made to historical data

Our seasonal adjustment package has an automatic procedure for dealing with outliers (observations that are far removed from the others in the series), which works well in most cases. However, in certain circumstances we need to deal with them explicitly. This is done by a prior adjustment.

A prior adjustment was made to the March 2008 and December 2012 quarters. This was made to male and female series, including full-time and part-time employment, and hours worked.

In these quarters we observed an unusually high level of transitions of people out of employment. This was particularly the case where individuals had been employed in the previous quarter and were then employed again in the subsequent quarter. The level of this type of behaviour was only observed in the March 2008, 2009, and December 2012 quarters.

Two of these quarters coincide with the Survey of Working Life in the March 2008 and December 2012 quarters, where we asked people who were employed additional questions to the standard HLFS about their working lives.

The size of the permanent prior adjustment was chosen by our seasonal adjustment programme with input into which quarters require the adjustment. The permanent prior adjustment improves the quality of, and coherence between, the trend series and seasonally adjusted series. Previously, the trend series had identified the December 2012 quarter observations for female employment and not in the labour force.

After further investigation, the permanent prior adjustment was also made to the total usual hours series.

Quality of seasonal adjustment

We monitor our data to make sure that our seasonal adjustment is robust.

The X-13ARIMA-SEATS programme is highly customisable and can produce a wide variety of possible adjustments for any particular input series. Consequently, X-13ARIMA-SEATS produces a number of diagnostics that are useful in assessing the quality of our chosen adjustment.

The following table provides a selection of diagnostics. The reference value indicates our desired value for each. Most are acceptable, though there is evidence of a changing seasonal pattern for the number of males who are employed and unemployed, and females who are not in the labour force.

More detailed information about seasonal adjustment in the HLFS is available on request:

 Seasonal adjustment diagnostics


  Reference value Male employed Female employed Male unemployed Female unemployed Male not in labour force Female not in labour force
Test for seasonality <0.10  0.00 0.00  0.00  0.00  0.00  0.00
Test for moving seasonality  >0.10   0.09 0.65 0.06  0.22  0.34   0.04 
Period until trend dominates  <3
Trend contribution to change  <20 33.39  38.80 45.19 14.59  14.06  19.29 
Seasonal contribution to change  >50 58.17  43.81  34.66  67.40 73.45 55.36 
Irregular contribution to change  <20   8.29  16.65  20.15  18.01  12.23  24.25 
Quality statistic  <1 0.38   0.50  0.90   0.72   0.53   0.84 
Evaluation of new series for seasonality

We have added several new series to the HLFS since the September 2013 quarter. In the September 2014 quarter we reviewed several of these series for seasonal patterns and for the possibility of publishing seasonally adjusted series.

We also reviewed the group of series that examine ethnicity by labour force status, and key seasonally adjusted series that are monitored every quarter.

We evaluated the following groups of series for evidence of seasonality and suitability for adjusting:

  • ethnicity by labour force status
  • age by labour force status (10-year age groups)
  • age by labour force status (new 5-year age groups: 65–69 and 70+).

Seasonal adjustment removes the seasonal pattern in a series to aid quarterly comparisons. A seasonal pattern is where a series consistently behaves in a particular manner for a given quarter. For example, a seasonal pattern in the employed series could be a ‘spike’ every December quarter; every year we see more individuals employed in the December quarter. It is possible that over time seasonal patterns cease (eg no spikes in the number employed in the December quarter) or seasonal patterns change (eg spike in the number employed is now in the March quarter).

See Seasonal adjustment in Statistics New Zealand for more information.

The review concluded there was not enough evidence to support seasonal adjustment of all series within the group 'age by labour force status' and 'ethnicity by labour force status'. However, we did find evidence of seasonality of series for younger age bands. We will consider how we might publish such statistics in the future.

We assessed the quality metrics for the following key series, to determine if there was consistent evidence between June 2013 and June 2014 of a cessation or change in the seasonal pattern:.

Female  Male Total both sexes 
Female employed Male employed Full-time
Female part-time employed  Male part-time employed Part-time
Female full-time employed  Male full-time employed Unemployed
Female unemployed  Male unemployed Total actual hours
Female not in the labour force  Male not in the labour force Total usual hours

There was consistent evidence of a change in the seasonal pattern for male full-time employed and total usual hours. There was some evidence to suggest a weak seasonal pattern for the male part-time employed and male unemployed series. All other key series presented evidence of an identifiable and stable seasonal pattern. Despite some series showing signs of a change in the seasonal pattern and weak seasonality, there was no strong evidence to suggest that seasonal adjustment should be stopped for any of the key series. We will continue to monitor these series every quarter.


During the seasonal adjustment process, X-13ARIMA-SEATS gives less weight to the irregular component. Specifically, if the estimated irregular component at a point in time is sufficiently large compared with the standard deviation of the irregular component as a whole, then the irregular component at that point can be downweighted or removed completely and re-estimated. Such observations are referred to as partial and zero-outliers, respectively. In practice, downweighting outliers does little to seasonally adjusted data, but the effect of the outliers on the trend series is generally reduced. However, if an outlier ceases to be an outlier as more data becomes available, then significant revisions to the trend series become possible.

No outliers are present over the last four quarters of data.

Suppression of data

We suppress cells with estimates of less than 1,000. They appear as ‘S’ in the tables. These estimates are subject to sampling errors too great for most practical purposes.

Rounding procedures

We round figures presented in this release. Figures are rounded to the nearest hundred or the nearest thousand for seasonally adjusted and trend estimates. This may result in a total disagreeing slightly with the sum of the individual items as shown in the table. Where figures are rounded the unit is shown as (000) if it is thousands.

We calculate any quarterly and annual changes for figures on unrounded numbers. However quarterly and annual percentage point changes for rates are done on rounded rates.

How labour force statistics are classified

The HLFS release includes specific statistics about industry, occupation, study, ethnicity, and region. This section defines what we measure for each of these statistics.

Industry statistics

Since the September 2009 quarter, our industry statistics are based on the Australian and New Zealand Standard Industrial Classification 2006 (ANZSIC06), the latest edition of the classification. When ANZSIC06 was introduced, we developed the New Zealand Standard Industrial Output Categories (NZSIOC). Classifying industries using NZSIOC helps to standardise outputs. Industry outputs defined using ANZSIC06 are not comparable with those based on ANZSIC96, the version used before the September 2009 quarter.

Implementing ANZSIC 2006 in the Household Labour Force Survey has more information.

Occupation statistics

Since the September 2009 quarter, we have used the Australian and New Zealand Standard Classification of Occupations (ANZSCO) to classify occupation data in the HLFS. ANZSCO is a harmonised classification developed by Statistics NZ, the Australian Bureau of Statistics, and the Australian Department of Employment and Workplace Relations, for use in both Australia and New Zealand. Occupation data was previously based on the New Zealand Standard Classification of Occupations 1999 (NZSCO99). The occupation data is available on Infoshare.

Implementing ANZSCO in the Household Labour Force Survey has more information.

Māori benchmarks

Before April 2009, we did not benchmark the Māori working-age population to population estimates. This, along with other sample design restrictions, caused a high degree of volatility in Māori statistics in the HLFS. Movements in the working-age population estimates of certain ethnic groups, such as Māori, may reflect this volatility rather than a real change in the estimated ethnic demographic. Including Māori benchmarks in the working-age population mitigates the known undercount of Māori in the HLFS and also results in smoother time series for Māori statistics. However, introducing the Māori population benchmarks does not necessarily translate to improved estimates for non-Māori ethnic groups.

Household statistics

We derive a household's labour force status by looking at the labour force status of household members aged 18–64 years. For example, if a couple is living by themselves and one is aged 64 years and the other is aged 65 years, this couple will be assigned to the 'All employed' or 'None employed' category, depending on the labour force status of the 64-year-old.

We exclude households that have no members aged 18–64 years from this analysis. The household categories incorporate the concept of dependent children rather than just children. A child is a person of any age who usually resides with at least one parent (natural, step, adopted, or foster) and who does not usually reside with a partner or children of his or her own. We define a dependent child as one under the age of 18 years and not in full-time employment.

Updated regional classification

In November 2010, the new Auckland territorial authority replaced the existing Rodney district, North Shore city, Auckland city, Waitakere city, Manukau city, Papakura district, and part of Franklin district councils. This resulted in a minor change in the boundary between the Auckland and Waikato regions.

From the June 2011 quarter, we produce the statistics in the HLFS release using the new boundaries, and backcast for the March 2011 quarter. The new boundaries do not significantly affect measures from the HLFS.

Total response ethnicity

From the December 2011 quarter, the HLFS publishes ethnicity data using the total response ethnicity output in the information release. Using this method, we count people who reported that they belonged to more than one ethnic group once in each group reported. This means that the total number of responses for all ethnic groups can be greater than the total number of people who stated their ethnicities.

Comparability with other datasets

Comparing our labour market statistics has more information on how the HLFS compares with the other labour market statistics we produce. This web page explains which measures of employment are included in each of our employment releases, and the timings and coverage of each release.

A Guide to Unemployment Statistics has more information on comparing the HLFS with other datasets on unemployment. This web page explains which measures of unemployment are included in the HLFS, jobseeker support – work ready, and the job seekers register. It also includes information on the timings, coverage, and different purposes of each of these measures.

HLFS comparable series 

The HLFS and the Quarterly Employment Survey (QES) are two different measures of employment and hours worked. The HLFS measures the number of employed people and the number of hours they usually work from New Zealand households; the QES measures the number of jobs and paid hours from New Zealand businesses. The HLFS comparable series removes major differences between HLFS and QES, yet does not make adjustments for all differences. This provides an HLFS series that is more comparable with QES.

In the June 2014 quarter, we made two changes to improve the seasonally adjusted HLFS comparable series.

  • We improved the methodology used to backcast the ANZSIC06 industry classifications. This provides more stable and robust industry estimates before the June 2009 quarter.
  • We excluded two further industry groups not covered in the QES (industry groups O7552 and L6711). This improves comparability of the HLFS and QES estimates. 

The HLFS comparable series removes the following categories from the HLFS, which are not collected by the QES:

  • self-employment
  • individuals who work without pay in a family business
  • the following industry groups:
    A01 Agriculture
    A02 Aquaculture
    A04 Fishing, hunting, and trapping
    A052 Agriculture and fishing support services
    L6711 Residential property operators
    O7552 Foreign government representation
    S96 Households employing staff
    T Not specified

The table below compares the annual percentage change of each survey's employment, and hours worked measures for recent quarters, in seasonally adjusted terms.

  Annual percentage change in employment Annual percentage change in hours

 Year to

HLFS comparable series people employed 

QES filled jobs

HLFS comparable series usual hours

QES hours paid

 Sep 2013 4.3  1.9  5.2 2.8
 Dec 2013 3.7  1.9  4.0 2.5
 Mar 2014 2.6  2.6  2.3 3.8
 Jun 2014  4.2   2.3  5.0 3.6
 Sep 2014 2.1  3.0  2.1 2.6

In the year to the September 2014 quarter, the HLFS comparable series reports higher growth in employment and hours than the QES does.

International comparability of the labour force participation rate and the employment rate

Several alternative definitions of labour force participation rate and employment rate are used by other organisations and countries; they differ in the age of the working-age population and the inclusion of military personnel. A common definition is to restrict the labour force and working-age population to the 15–64-year age group, particularly in countries with a compulsory retirement age. Generally, this definition leads to a higher labour force participation rate and employment rate.

Using this definition for the New Zealand HLFS in the September 2014 quarter gives a surveyed figure of 78.7 percent (labour force participation rate) and 74.2 percent (employment rate).

Interpreting the data

Information releases contain seasonally adjusted, trend, and survey statistics for the latest quarter. These statistics are averages for the three-month period and do not apply to any specific point in time. We identify data sourced from the seasonally adjusted series and trend series as such in the table or section headings. All other data, in the commentary or in tables, are sourced from the original survey series and are unadjusted.

Timing of published data

The HLFS is published within six weeks after the end of the quarter's reference period.


Only people authorised by the Statistics Act 1975 are allowed to see your individual information, and they must use it only for statistical purposes. Your information is combined with similar information from other people or households to prepare summary statistics.

More information

See more information about the Household Labour Force Survey

Statistics in this release have been produced in accordance with the Official Statistics System principles and protocols for producers of Tier 1 statistics for quality. They conform to the Statistics NZ Methodological Standard for Reporting of Data Quality.


While all care and diligence has been used in processing, analysing, and extracting data and information in this publication, Statistics NZ gives no warranty it is error-free and will not be liable for any loss or damage suffered by the use directly, or indirectly, of the information in this publication.


Our information releases are delivered electronically by third parties. Delivery may be delayed by circumstances outside our control. Statistics NZ accept responsibility for any such delay.

Crown copyright©

Creative Commons logo.
This work is licensed under the Creative Commons Attribution 3.0 New Zealand licence. You are free to copy, distribute, and adapt the work, as long as you attribute the work to Statistics NZ and abide by the other licence terms. Please note you may not use any departmental or governmental emblem, logo, or coat of arms in any way that infringes any provision of the Flags, Emblems, and Names Protection Act 1981. Use the wording 'Statistics New Zealand' in your attribution, not the Statistics NZ logo.

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