Stats NZ has a new website.

For new releases go to

www.stats.govt.nz

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+
Productivity Statistics: 1978−2009
Embargoed until 10:45am  –  16 March 2010
Commentary

Productivity is a measure of how efficiently inputs are being used within the economy to produce outputs. Productivity is commonly defined as a ratio of a volume measure of output to a volume measure of input. Growth in productivity means that a nation can produce more output from the same amount of input. Productivity growth is an important contributing factor to a nation’s long-term material standard of living.

Annual productivity series for the measured sector from 1978–2009 have been released in this publication. In 2007, the latest year for which current price industry value added data are available, the measured sector covered approximately 74 percent of the economy. It excludes the following industries: government administration and defence, health, education, ownership of owner-occupied dwellings, and property services. From 1996 onwards, business services, and personal and other community services are included in the measured sector. See 'Industry coverage – the measured sector' in 'Technical notes' for more detail on the coverage of the measured sector and former measured sector.

The additional tables 1.1–1.8 include data for the former measured sector from 1978–2009. This data maintains comparability with the previously released time series. The former measured sector is the most appropriate sector for comparing with Australian productivity data, which has recently transferred to a new industry classification system. See the ‘Comparison with Australia’ section for further detail.

Productivity estimates are presented on an annual basis and across business cycles. A business cycle is defined as 'peak to peak'. Estimating growth over business cycles is preferable as it accounts for changes in capacity utilisation rates over the period of a business cycle. New to this release is the identification of 2006 as a peak, which means 2000–06 is a complete cycle. Please note that the latest period (2006–09) is not a complete cycle. Caution is advised when comparing the latest period with other cycles. For more information, see the 'Estimating growth cycles' section in 'Technical notes'.

Unless otherwise stated, all references to average movements are annual geometric mean movements relating to the measured sector.

Labour productivity

 Graph, Measured sector labour and output indexes, year ended March, 1978–2009.

Labour productivity is measured as a ratio of output to labour input. In the year ended March 2009, labour productivity decreased by 1.5 percent, due to a fall in both output (down 2.2 percent) and labour input (down 0.7 percent).

The following table presents the average annual growth in labour productivity for the growth cycles identified within the series.

Labour Productivity Average Annual Growth Rates(1)(2)
Year ended 31 March
Cycle Output Labour input Labour productivity
Percent(3)
1978–82 2.1 0.4 1.7
1982–85 3.4 1.8 1.5
1985–90 0.7 -2.1 2.9
1990–97 3.3 0.7 2.6
1997–2000 3.0 -0.1 3.1
2000–06 3.8 2.4 1.3
2006–09 0.5 0.8 -0.3
1978–2009 2.5 0.6 1.9
(1) The average annual growth rate values do not include the movement for the first year of each cycle, eg the 1978–82 average annual growth rate does not include the movement for 1978.
(2) Business services, and personal and other community services are included in the measured sector from 1996 onwards.
(3) Percentage changes are calculated on unrounded index numbers.


A fall in labour productivity of 1.5 percent for the March 2009 year is below the average annual growth rate for the 2006–09 period of -0.3 percent. The growth rate over this latest period is lower than any of the previous cycles and is the only cycle where labour productivity is negative. However, caution should be exercised in this comparison as the 2006–09 period is not a complete cycle.

Labour input growth averaged 0.8 percent annually over the 2006–09 period. In the year to March 2009, labour input fell 0.7 percent marking the first decrease in this figure in 10 years. The downward movement in labour input growth in the year to March 2009 is consistent with the changes in the labour market during this time.

The weakening of the labour market was highlighted by the unemployment rate increasing to 5.0 percent for March 2009 quarter (up from 3.8 percent for the March 2008 quarter). The unemployment rate is for the whole economy, as an unemployment rate that covers only the measured sector is not available. Indications of a weak labour market are also evident in the New Zealand Institute of Economic Research's Quarterly Survey of Business Opinion (April 2009) which shows that a net 42 percent of respondents thought it was easier to find skilled or technical staff in March 2009 compared with three months prior. The comparable figure for unskilled staff was 63 percent. Both figures are the highest for more than 30 years.

The decline in labour productivity during the 2006–09 cycle has not significantly affected the long-term average growth rate for labour productivity, which remains positive.

Capital productivity

 Graph, Measured sector capital and output indexes, year ended March, 1978–2009.

Capital productivity is measured as a ratio of output to capital input. In the year to March 2009, capital productivity fell 5.3 percent, due to capital inputs increasing by 3.3 percent and output decreasing by 2.2 percent.

The following table presents the growth in capital productivity for the growth cycles identified within the series.

Capital Productivity Average Annual Growth Rates(1)(2)
Year ended 31 March
Cycle Output Capital input Capital productivity
Percent(3)
1978–82 2.1 1.4 0.8
1982–85 3.4 5.5 -2.0
1985–90 0.7 4.3 -3.5
1990–97 3.3 1.9 1.3
1997–2000 3.0 2.6 0.4
2000–06 3.8 3.9 -0.1
2006–09 0.5 3.7 -3.1
1978–2009 2.5 3.2 -0.7
(1) The average annual growth rate values do not include the movement for the first year of each cycle, eg the 1978–82 average annual growth rate does not include the movement for 1978.
(2) Business services, and personal and other community services are included in the measured sector from 1996 onwards.
(3) Percentage changes are calculated on unrounded index numbers.


The decline in capital productivity over the 2006–09 period can be explained by the annual increase in capital input of 3.7 percent, relative to the small rise in output of 0.5 percent annually. In 2009, the capital input increase was driven by strong investment in IT-related assets. Although capital productivity declined sharply in 2009, it is possible utilisation of capital was at a low point. The productivity statistics do not adjust for changes in capacity utilisation, as a capital asset is assumed to be used at a constant rate over its life. Under conditions where utlilisation of capital is lower than average, growth in capital inputs may be artificially high and therefore growth in capital productivity may be artificially low. To partly counter this scenario, it is recommended that productivity is analysed using business cycles. A decline in capital productivity is seen in four of the seven cycles. Caution should be exercised in this comparison as the 2006–09 period is not a complete cycle.

Over the entire time series (1978–2009), capital productivity fell 0.7 percent on an average annual basis. This was due to annual capital input growth of 3.2 percent increasing by more than the annual growth in output of 2.5 percent.

Capital deepening is defined as the relative increase in capital inputs to that of labour inputs, or a rise in the capital-to-labour ratio. In the latest period, 2006–09, the capital-to-labour ratio has continued to increase with an annual growth rate of 2.8 percent, due to 3.7 percent annual growth in capital input coupled with 0.8 percent growth in labour input. The New Zealand economy has experienced capital deepening during the last 31 years. The annual growth in the capital-to-labour ratio was 2.6 percent over the entire 1978–2009 series. This was due to capital input growth of 3.2 percent, compared with 0.6 percent growth in labour input. The capital deepening figures only provide information on the growth rate in the capital-to-labour ratio. They do not shed insight into the level of capital per worker in the economy.

Multifactor productivity

 Graph, Measured sector input, output, and productivity indexes, year ended March, 1978–2009.

Multifactor productivity (MFP) is measured as a ratio of output to total inputs. It can also be defined as growth that cannot be attributed to capital or labour, such as technological change or improvements in knowledge, methods, and processes. In the year to March 2009, MFP declined 3.1 percent, due to total inputs increasing (up 1.0 percent) while output decreased (down 2.2 percent).

The following table presents the annual average growth in MFP within the growth cycles identified for the series.

Multifactor Productivity Average Annual Growth Rates(1)(2)
Year ended 31 March
Cycle Output Total inputs Multifactor productivity
Percent(3)
1978–82 2.1 0.7 1.4
1982–85 3.4 3.2 0.2
1985–90 0.7 0.3 0.4
1990–97 3.3 1.1 2.1
1997–2000 3.0 0.9 2.0
2000–06 3.8 3.1 0.7
2006–09 0.5 2.0 -1.5
1978–2009 2.5 1.6 0.9
(1) The average annual growth rate values do not include the movement for the first year of each cycle, eg the 1978–82 average annual growth rate does not include the movement for 1978.
(2) Business services, and personal and other community services are included in the measured sector from 1996 onwards.
(3) Percentage changes are calculated on unrounded index numbers.


From 2006–09, MFP has declined for the first time across a cycle in the series, with an average decrease of 1.5 percent annually. Output rose 0.5 percent on an average annual basis, while total inputs rose 2.0 percent. The main contributor to the growth in total inputs was capital input. Caution should be exercised in this comparison as the 2006–09 period is not a complete cycle.

The average annual increase of 0.9 percent in MFP over the entire 1978–2009 series was due to output (up 2.5 percent) rising more than input growth of 1.6 percent.

Growth accounting for real GDP

Growth accounting examines how much of the economy’s growth in output can be explained by the growth of combined inputs. In particular, growth in real GDP (output) can arise from three different sources: an increase in labour input, an increase in capital input, or an increase in MFP. The following graph presents growth in output between 1978 and 2009 for the growth cycles identified in the series.

 Graph, Contribution to measured sector real GDP growth, average percentage change over period, 1978–2009.

In the 2009 March year, output fell by 2.2 percent. This decline was driven by MFP (down 3.1 percent), and to a lesser extent by labour input (contributing -0.4 percent to output growth). These were partly offset by capital input, which contributed 1.3 percent to GDP.

The following table presents the annual average growth in output and its contributing factors for the growth cycles identified within the series. 
 

Contribution to Measured Sector Real GDP Growth
Average annual growth rates (1)(2)
Year ended 31 March
Cycle Real GDP Contribution of
capital input(3)
Contribution of
labour input(4)
Multifactor productivity
Percent(5)
1978–82 2.1 0.5 0.2 1.4
1982–85 3.4 2.1 1.1 0.2
1985–90 0.7 1.7 -1.3 0.4
1990–97 3.3 0.8 0.4 2.1
1997–2000 3.0 1.0 -0.1 2.0
2000–06 3.8 1.6 1.4 0.7
2006–09 0.5 1.5 0.5 -1.5
1978–2009 2.5 1.3 0.3 0.9
(1) The average annual growth rate values do not include the movement for the first year of each cycle, eg the 1978–82 average annual growth rate does not include the movement for 1978.
(2) Business services, and personal and other community services are included in the measured sector from 1996 onwards.
(3) Contribution of capital input is equal to the growth rate in capital input weighted by capital's share of total income.
(4) Contribution of labour input is equal to the growth rate of labour input weighted by labour's share of total income.
(5) Percentage changes are calculated on unrounded index numbers.


In the current cycle (2006–09) the average GDP growth rate is the lowest since the series began. The key industries which have been driving GDP down over this time are manufacturing (in 2007 and 2009), agriculture (2008), and construction (2009).

Over the entire 1978–2009 series, output growth averaged 2.5 percent. Capital input was the largest contributor, averaging 1.3 percent annually. Labour input contributed 0.3 percent to this rise in output and MFP contributed 0.9 percent.

Growth accounting for labour productivity

As with growth in real GDP (output), growth in labour productivity can be broken down into components. In particular, a change in labour productivity can come from two possible sources: a change in the weighted capital-to-labour ratio (that is, capital deepening or capital shallowing) and a change in MFP. The following graph presents the contributions to labour productivity growth between 1978 and 2009 for the growth cycles identified in the series.

 Graph, Contribution to measured sector labour productivity growth, average percentage change over period, 1978–2009.

In the year to March 2009, labour productivity declined 1.5 percent. The annual contribution from capital deepening was 1.6 percent, which partly offset a decline of 3.1 percent in MFP.

The following table presents the annual average growth in labour productivity and its contributing factors for the growth cycles identified for the series.

Contribution to Measured Sector Labour Productivity Growth
Average annual growth rates(1)(2)
Year ended 31 March

Cycle Labour productivity Contribution of capital deepening(3) Multifactor productivity

Percent (4)

1978–82 1.7 0.3 1.4
1982–85 1.5 1.4 0.2
1985–90 2.9 2.5 0.4
1990–97 2.6 0.5 2.1
1997–2000 3.1 1.1 2.0
2000-06 1.3 0.6 0.7
2006–09 -0.3 1.2 -1.5
1978–2009 1.9 1.0 0.9
(1) The average annual growth rate values do not include the movement for the first year of each cycle, eg the 1978–82 average annual growth rate does not include the movement for 1978.
(2) Business services, and personal and other services are included in the measured sector from 1996 onwards.
(3) Contribution of capital deepening is equal to the growth rate in the capital-to-labour ratio weighted by capital's share of total income.
(4) Percentage changes are calculated on unrounded index numbers.


Labour productivity over the latest 2006–09 period declined by 0.3 percent annually. This was driven by a decrease of 1.5 percent in MFP growth, offset by an increase in the contribution of capital deepening (up 1.2 percent). Capital deepening has shown fairly strong growth over the period, while MFP growth has largely been negative.

Over the entire 1978–2009 series, the average annual contribution to labour productivity growth from capital deepening was 1.0 percent. The average contribution of MFP growth was 0.9 percent on an annual basis. Labour productivity growth averaged 1.9 percent annually.

Composition-adjusted productivity

Within the suite of productivity measures, there are three labour input series. These are: the unweighted, unindexed measure of labour volume; the headline labour input index; and the composition-adjusted labour input index.

The composition-adjusted input series allows us to track changes in the skill level of the workforce over time. To do this, it needs to be compared with the unweighted labour volume series. The difference between these two series represents the change in the skill level – measured using qualification and experience proxies – of workers.

This composition-adjusted input series explicitly accounts for quality. It is generally considered to provide the most representative measure of labour input. Alternatively, the headline labour input series implicitly adjusts for quality. It does this by giving higher weight to industries with above-average wage rates.

In the year ended March 2009, composition-adjusted labour productivity decreased by 1.2 percent. This was due to adjusted labour input (down 1.0 percent) decreasing by less than the fall in output of 2.2 percent. Composition-adjusted multifactor productivity decreased by 2.9 percent, as adjusted total inputs rose 0.7 percent.

The table below shows the annual movements for each of the three labour input series. It begins in 1999, the first year in which composition-adjusted labour input movements are available.

Annual Percentage Change in Labour Input Series: 1999–2009

Year

Labour volume measure

Headline measure

Composition-adjusted measure

1999 -0.7 -0.9 -0.1
2000 1.4 0.8 1.9
2001 1.6 1.4 2.2
2002 2.9 3.0 2.9
2003 3.0 3.2 3.3
2004 3.1 3.0 3.8
2005 2.0 2.4 2.1
2006 1.4 1.7 1.6
2007 1.3 1.6 1.1
2008 1.4 1.5 2.2
2009 -0.8 -0.7 -1.0
Average 1999–2009(1) 1.5 1.5 1.8
(1) Calculated as the geometric mean growth rate across the period


The composition-adjusted series is clearly growing faster than either the unweighted labour volume series or the headline measure. As labour productivity is measured as residual output growth above labour input growth, the implication is that composition-adjusted productivity is growing slower than the other productivity measures.

Comparing the composition-adjusted series with the labour volume measure, the only years in which it grows slower are 2007 and 2009. In 2002 they have the same growth rate. In all other years, the labour input of higher-skilled workers, as measured using the education and experience proxies, is growing faster than that of lower-skilled workers. This implies that the average skill level of the workforce is increasing. 

Composition-adjusted productivity growth accounting for real GDP

The inclusion of a composition-adjusted labour input series allows for the decomposition of growth in real GDP (output) into capital input, hours paid, labour composition, and composition-adjusted MFP.

The following graph presents the contributions to annual growth in output for the adjusted series from 1998 to 2009. Labour composition changes have not been a major driver of increases in GDP over this time period.

 Graph, Contribution to measured sector real GDP growth, incorporating composition-adjusted labour input, annual percentage change 1999–2009.

In the year to March 2009, growth in output declined by 2.2 percent. This decline was driven by adjusted MFP, which fell 2.9 percent, and to a lesser extent by hours paid and labour composition (contributing -0.5 percent and -0.1 percent, respectively). Capital input contributed 1.3 percent to real GDP growth.

The contribution of labour composition to growth in output is positive in all years but 2002, 2007, and 2009. In all other years, highly-skilled labour is growing at a faster rate, and therefore the average skill level of workers is rising. Under an unadjusted labour input approach, this positive contribution of labour skill would remain part of the labour productivity and MFP residuals and is not explicitly shown. Explicitly accounting for the rising skill composition allows us to explain more of the residual growth in output.

Revisions

Updates in data sources and ongoing methodology improvements have caused a number of revisions to the previously published productivity series. Please refer to tables 7, 8, 1.7, and 1.8 in the downloadable Excel spreadsheets for the magnitude and direction of these revisions.

Regular revisions (due to updates in data sources) have arisen from the following:

  • revised constant price GDP data, feeding into the output series
  • revised current price national accounts data, with current price industry data now available for 2006 and 2007, feeding into the industry income-based weights
  • revised current and constant price productive capital stock data for selected assets and industries for March years 2006, 2007, and 2008, feeding into the capital input series
  • the addition of linked employer-employee data (LEED) to replace survey-based data for working proprietor counts for the March 2008 year, and employee counts for the March 2008 quarter, feeding into the labour input series.

Revisions resulting from improved methodology are:

  • an improvement to the methodology for linking LEED data in 2000 to create a continuous labour series and a change to the use of LEED to make it more conceptually consistent with GDP data from the National Accounts
  • the addition of business demography data for 1987, for most industries
  • the addition of inventories to the set of assets within the capital input series, from 1987 onwards
  • a change to the methodology for timber and livestock assets within the capital input series.

Due to the amount of provisional data that is used in productivity calculation for the most recent years, the last four years of the series are released with provisional status. While there have been revisions to some annual movements, the underlying trend of the productivity series has remained unchanged.


Comparison with Australia

Official New Zealand productivity data can be benchmarked against official Australian numbers. The use of the same industrial classification (Australian and New Zealand Standard Industrial Classification 1996, or ANZSIC96) has made such comparisons more valid since Statistics NZ began publishing productivity statistics.

However, the Australian Bureau of Statistics (ABS) is now publishing their national accounts and productivity data under ANZSIC 2006. Given that Statistics NZ will not publish under ANZSIC 2006 until 2012, this has distorted the comparison.

Under ANZSIC96, the ABS market sector had identical industry coverage to the Statistics NZ former measured sector. To maintain some continuity for their users, the ABS is now publishing an ANZSIC06-based series which is as close as possible to their market sector, and therefore as close as possible to the Statistics NZ former measured sector. This series, known as MFP12, dates back to 1974 and is currently the best series to use when undertaking a trans-Tasman comparison. The ABS ANZSIC96-based market sector series can still be used for comparisons up to 2008, and is available on their website, www.abs.gov.au.

In 2007, both the ABS’ MFP12 series and the Statistics NZ former measured sector covered 63 percent of the economy. The table below shows which industries are included in each series.

Australia’s MFP12 and New Zealand’s Former Measured Sector

Australia (ANZSIC 2006-based) New Zealand (ANZSIC 1996-based)
A – Agriculture, forestry, and fishing A – Agriculture, forestry, and fishing
B – Mining B – Mining
C – Manufacturing C – Manufacturing
D – Electricity, gas, water, and waste wervices D – Electricity, gas, and water
E – Construction E – Construction
F – Wholesale trade F – Wholesale trade
G – Retail trade G – Retail trade
H – Accommodation and food services H – Accommodation, cafes, and restaurants
I – Transport, postal, and warehousing I – Transport and storage
J – Information media, and telecommunications J – Communication services
K – Financial and insurance services K – Finance and insurance
R – Arts and recreation services P – Cultural and recreational services

 
The table below compares the output, productivity, and input series for Australia and New Zealand, based on the MFP12 and former measured sector series, respectively. The Statistics NZ time series begins in 1978, so the comparison is over 31 years. A cycle comparison cannot be undertaken, due to each country having slightly different peaks in their cycles.


Comparison of Australia and New Zealand Productivity Data
Australia's MFP12 and New Zealand's former measured sector
Average annual growth rates: 1978–2009

Variable Australia New Zealand
Output 3.1 2.4
Labour productivity 2.0 2.1
Capital productivity -0.7 -0.6
Multifactor productivity 0.8 1.0
Labour input 1.1 0.3
Capital input 3.9 3.0
Total inputs 2.3 1.4
Capital-to-labour ratio 2.8 2.7


Based on the 63 percent of the economy covered by these statistics, Australia’s and New Zealand’s productivity performances have been very similar over the 1978–2009 period. New Zealand is slightly ahead in all three measures of productivity growth. Australia’s output growth is significantly higher at 3.1 percent annually compared with New Zealand’s 2.4 percent. However, this additional output growth is fully accounted for by input growth. Both Australia and New Zealand have experienced similar growth in capital deepening over the 31-year time span, with Australia nudging slightly ahead.

 

For technical information contact:
Brendan Mai or Nicholas Warmke
Wellington 04 931 4600
Email: productivity@stats.govt.nz

Next release ...

Productivity Statistics: 1978–2010 is scheduled to be released in March 2011.

Related releases ...

Exploring the Possibility of Measuring Government Sector Productivity in New Zealand: A feasibility study will be released on 18 March 2010.

Industry Level Productivity Statistics: 1978–2008 is scheduled to be released in June 2010.

  • 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+