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+
Frequently asked questions about productivity – Labour

What is labour productivity?

Labour productivity growth reflects the change in the amount of output per hour paid. An increase in labour productivity means that an hour of paid work is producing more output than in the previous year, or that the same amount of output is being produced for fewer hours paid.

Why have so many different data sources been used?

The OECD guidelines suggest making use of the comparative advantages of different statistical sources for more robust results. The primary data sources are the QES, BDD, and LEED (from 2000 onwards). The first two sources are establishment-based, and are supplemented with census and HLFS data for working proprietors and for industries excluded from the QES. LEED is an administrative data source which is created using data from the Statistics NZ Business Frame and administrative taxation sources.

Different data sources are used to measure labour input as no one data source is suitable on its own over a long time series. A combination of sources is therefore the best approach. While acknowledging that there are conceptual differences between the data sources, these are not significant enough to prohibit their use.

Aren't most of these data sources available on a quarterly basis?

Yes. This implies there is potential to construct a quarterly labour productivity series. Statistics NZ has considered this, but there are still no formal plans to release an 'official' quarterly labour productivity series. A quarterly labour volume series is available on request.

Why has ‘hours paid’ been used as the input measure instead of ‘hours worked’?

Conceptually, for measuring productivity, hours worked is a better measure of labour input than hours paid. Hours paid is the number of ordinary and overtime hours for which an employee is paid, and does not cover unpaid overtime which is part of labour input. It also includes holiday pay. Hours paid is the only input measure available from the QES.

Statistics NZ has chosen to use hours paid because:

  • industry estimates of hours paid are more reliable. Robust industry data and good alignment with output and capital data are important for producing industry measures of productivity.  At the industry level, the hours paid measure is more robust than the hours worked measure. Given that the methodology used to calculate the labour volume index is based on the aggregation of industry level data, it is desirable to have a good alignment of industry inputs and outputs.
  • hours paid provides a longer historical time series. Productivity analyses require a long time series to examine productivity trends over business cycles. Hours paid data is available back to the early 1970s.

Why does the labour volume series show lower growth than the measured sector HLFS actual/usual hours series?

The relationship between the two series has been very stable since 1988. However, they differ from 1992–95. During this period, the HLFS grew significantly faster than the labour volume series. This may be due to several factors including:

  • A sample redesign of the HLFS was undertaken and the new sample was introduced progressively over the four quarters from December 1993 to September 1994. During this time the new sample showed a consistently higher level of employment.
  • The changing structure of the economy. Around 1993, the economy was coming out of a recession and was experiencing a period of strong growth. There were also major changes to the legislative framework of the labour market in the form of the Employment Contracts Act.

Why does the measured sector indexed labour series (labour input series) show different growth than the unindexed series (the labour volume series)?

The labour input series is created as a chain-linked Törnqvist index, using industry income shares as weights, partly to account for differences in the 'quality' of labour. Those industries with above average hourly compensation will be allocated a higher weight in the labour input index than in the labour volume series.

The indexed (labour input) series shows lower growth than the unindexed (labour volume) series over time because much of the growth in measured sector labour input has been in lower paid industries, such as accommodation, cafes, and restaurants; and retail trade.

Why is LEED the primary data source for employee and working proprietor counts? And why has it been used only from 2000?

The LEED dataset is created by linking a longitudinal dataset from the Statistics NZ Business Frame with longitudinal data from administrative taxation sources. This dataset is available only from the June 1999 quarter. LEED is considered to be the best available data source for measuring labour counts because:

  • The employee counts data obtained from LEED is available on a monthly basis, allowing us to better capture seasonality changes in labour volume.
  • LEED counts are not based on survey information which reduces the effect of sample error on the series.
  • It provides information on secondary jobs for industries outside of the scope of the QES. These jobs were previously excluded from the series.
  • LEED provides annual benchmarks for working proprietors, removing the need to extrapolate census benchmarks for up to five years.
  • Working proprietors who pay themselves a salary can now be identified more accurately.

How does the LEED data used within productivity estimates compare with published LEED data?

LEED data used within productivity estimates are at the kind-of-activity unit (KAU) level, while published LEED data are based on geographic (GEO) unit levels. A KAU is the part of an enterprise producing goods and services, with a single set of accounting records. GEO unit levels are separate operating units engaged in one or predominantly one kind of economic activity, from a single site.

As a KAU can include more than one GEO unit, industry totals may differ when summarised to these levels. As GDP is calculated on a KAU basis, and productivity relates GDP growth to input growth, the inputs need to be measured as consistently as possible to GDP.

Other differences between the LEED measures are that some working proprietor values from the monthly LEED employee counts are extracted to prevent double counting.

Why did you decide to rate forward the labour series for the latest year when it has been 12 months since your reference period?

The labour series was rated forward as our primary data source (LEED) was not available for the whole period. LEED data is based on Inland Revenue (IRD) forms, which are subject to late filing from taxpayers. This is why both the latest quarter of employment data and the latest year of working proprietor data are rated forward. LEED working proprietor data for the year ended March is normally released approximately 18 months later.

How did you rate the series forward?

Survey-based information from the HLFS and QES was used to rate forward the series.

  • 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+