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+
Gross Domestic Product: September 2012 quarter
Embargoed until 10:45am  –  20 December 2012
Revisions

Summary of revisions

We incorporated several revisions in this release, with the following table showing the industries affected. Each major revision is discussed in a separate section below.  

 
Industry
 Revision
Reconciliation to annuals Financial intermediation services indirectly measured  Owner-occupied dwellings  Other methodology changes 
Agriculture, forestry, and fishing

 x

 x

    
Mining

 x

 x

   
Manufacturing

 x

 x

   
Electricity, gas, water, and waste services

 x

 x

   
Construction

 x

 x

   
Wholesale trade

 x

 x

   
Retail trade and accommodation

 x

 x

   
Transport, postal, and warehousing

 x

 x

   
Information media and telecommunications

 x

 x

   
Financial and insurance services

 x

 x

   
Rental, hiring, and real estate services

 x

 x

 x

 
Prof, scientific, technical, admin, and support

 x

 x

   
Public administration and safety

 x

 x

   
Education and training

 x

 x

 

 x 

Health care and social assistance

 x

 x

 

 x

Arts, recreation, and other services

 x

 x

   
Symbol: x – industry affected by revision

Quarterly series reconciled to new annual balanced data

Reconciling quarterly constant price series to balanced annual series is an integral process in compiling national accounts. Annual data is balanced in a supply and use framework, so that the supply of all goods and services in the economy (goods and services produced by the New Zealand economy and imports from overseas) is equal to the use of these goods and services (final consumption by households and government, investment and exports). The annual supply and use process uses much more data and more detail than is available to compile quarterly GDP. The annual balancing process is done in current prices, with the resulting values converted into constant price series using various internationally accepted practices. This reconciliation process resolves the problem of combining high-frequency data (quarterly) with a series of low-frequency data (annual).

The reconciliation process maintains the relevance and coherence of national accounts statistics. It is the main opportunity to enhance the quality of volume estimates and to align the production and expenditure measures of GDP.

The new annual data was incorporated into this release come from the National Accounts (Industry Benchmarks): Year ended March 2010 and National Accounts (Income and Expenditure): Year ended March 2012 published on 21 November 2012.

The revisions to QGDP as a result of incorporating new annual benchmarks and methodological changes are shown in the graphs below.  

Graph, Gross domestic product, revisions to levels, September 2006 to September 2012.

Graph, Gross domestic product, revisions to growth rates, September 2006 to September 2012.

Allocating financial intermediation services indirectly measured

Banks and similar institutions provide financial services. These services are paid for explicitly through fees, or implicitly through interest rates. The explicit fees (such as transaction and account fees) can be directly measured. However, this is not possible for the implicit fees associated with the financial intermediation (borrowing and lending) services banks provide. Banks provide an intermediary mechanism by accepting deposits from one party and lending these funds to another party. In doing so, the bank pays the depositor an interest rate that is reduced by a service fee, while charging the borrower a rate increased by a fee. These implicit fees represent charges for the financial intermediation services indirectly measured (FISIM).

Previously, bank service charges were included in the measurement of GDP as a notional industry that had a negative value. This treatment assumed that all bank service charges were paid by industries, therefore resulting in an understatement of GDP. The new treatment recognises that some FISIM is paid by final consumers such as households, reducing the amount paid by industries. This reduction has raised the production measure of GDP (by lowering intermediate consumption) and the expenditure measure of GDP (by increasing final demand). Previously, all bank service charges were assumed to have been paid by one notional industry.

This revised treatment is in line with the recommended international measurement standards defined in the System of National Accounts.

Implementing FISIM into the quarterly accounts depends on the method used to derive constant price value added for that industry. For industries that use volume extrapolation from the base year, the base year is lowered due to the additional FISIM charge, but the quarterly movements are kept the same. Therefore, implementing FISIM in these industries results in a change in levels only. For industries that use the double-deflation method, where constant price value added is measured by separately deflating outputs and inputs, the level of value added will be lower, and the annual growth rates may also change.

The volume measure of FISIM will reflect changes in the real value of loans and deposits. The changes in the real values are derived by deflating loan and deposit balances by a price measure associated with the type of expenditure. For example, household deposits are associated with household consumption expenditure and are deflated by the consumers price index (CPI – all groups series). Changes to banks' interest margins do not affect the volume measure of FISIM.

Another impact of replacing the bank service charge with FISIM is that the output of the financial sector has fallen. The scope of FISIM is narrower than the bank service charge and has replaced it as the output indicator for the finance industry. This change affects the level of output for this industry as well as its growth rates. The bank service charge was derived from the flows of all interest-bearing assets and liabilities (such as bonds, notes, and securities) while FISIM is derived only from loans and deposits.

See Financial intermediation services indirectly measured (FISIM) in the national accounts for more information about FISIM and how it is calculated.  

New method for deriving owner-occupied dwellings

In the national accounts, owner-occupied dwellings (OOD) is an industry that measures the services provided to people living in houses they own. The value of these OOD services is calculated as an imputed rent. In other words, the output of owner-occupied dwellings is valued at the estimated rent that a tenant would pay for the same accommodation.

OOD services are included in GDP because the ratio of owner-occupied to rented dwellings can change. Comparisons of the production and consumption of housing services over time could be distorted if no estimate was made for the value of own-account housing services. A secondary reason for including OOD is that home ownership rates can vary by country, so the output of housing can be compared across countries regardless of home ownership rates.

A new methodology was introduced for OOD. No changes were made to the current price annual method for OOD and to the quarterly indicator. The new methodology only affects the annual constant price measure of output, and subsequently flows through to the value added of this industry and through consumption of OOD rental services in household consumption expenditure.

The constant price annual method for OOD is a double indicator method – this means there are different indicators for gross output and intermediate consumption. The indicator for intermediate consumption is unchanged – this was and still is deflated current price intermediate consumption.

The new annual methodology for constant price OOD output is to deflate annual current price output to derive an annual constant price series. The deflator used is a specific producers price index for owner-occupied dwelling services that accounts for quality change. The new annual methodology is also reflected in the expenditure measure of GDP. Imputed rents for OOD services in annual household consumption expenditure equal the output as calculated on the production side of GDP.

The previous method of calculating the constant price output of OOD was volume extrapolation, using the number of dwellings from population estimates as an indicator. This methodology did not account for quality change. The number of dwellings is still being used as a quarterly indicator where reconciled annuals are not available.

The overall effect of the new quality adjusted method for annual chain-volume estimates of OOD is higher levels and growth rates for the series.

It is also worth noting that OOD is one of the industries heavily affected by the introduction of FISIM (see previous section above). Interest payments are significant in this industry. The introduction of FISIM increases the level of intermediate consumption for the OOD industry, which has the effect of reducing value added. The graph below shows the impact of these changes on the industry's total value added.  

Graph, owner-occupied dwellings, revisions, September 2006 to September 2012.

New estimates for education and training, and health care and social assistance

Improved methodologies for measuring the education and training, and health care and social assistance industries were also incorporated in this release. These new methodologies were previously signalled in Measuring government sector productivity in New Zealand: A feasibility study. The changes affect the annual chain-volume series to which the quarterly series are reconciled. There are no changes to the quarterly methods, or the annual current price measures for these industries.

Education and training

For education and training, improved annual benchmarks were introduced by further refining the output volume series for pre-school, schools, and tertiary education. For each of these series, the output measure is a cost-weighted index of equivalent full-time students (EFTS). This index is taken to represent the volume of teaching services provided by each of these types of institutions, and is used to extrapolate base year value added. The previous methodology derived an output index based on student numbers (rather than EFTS), and omitted certain types of pre-school and tertiary institutions (such as playcentres, te kohanga reo, and wānanga).

Health care and social assistance

Methodology improvements were introduced for the health care and social assistance industry. The indicator for government hospitals in this industry was broadened to cover all hospital events (inpatient, day patient, and emergency). These indicators better reflect the changes in the type of treatments for patients in the output of health care. Annual constant price value added is now calculated using a double indicator method, where intermediate consumption and gross output are independently deflated to calculate value added. 

The previous method for government hospitals was based on extrapolating base year value added by a volume index, mainly using data on inpatient discharges.

Revisions to value added for these industries, as a result of the new methodologies, are shown in the graphs below. 

Graph, Education and training, revisions, September 2006 to September 2012.

Graph, health care and social assistance revisions, September 2006 to September 2012.

An information paper with more detail on the measurement of value added in education and training, and health care and social assistance will be released on 25 January 2013.  

Revisions to the expenditure measure of GDP

Household consumption expenditure

Revisions to the household consumption expenditure series were due to:

  • reconciliation to new annual current price measures as published in National Accounts (Income and Expenditure): Year ended March 2012
  • introduction of FISIM (see section above) which has increased final expenditure by households
  • introduction of new OOD estimates (see section above) which increased the volume of services of housing consumed by households
  • revisions from the Retail Trade Survey: September 2012 quarter
  • a review of the household consumption expenditure system also led to revisions due to improved proportional splitting, and improvements to how the quarterly series reconciles to annual data.

Central government final consumption expenditure

The system used to compile central government expenditure was upgraded. This resulted in revisions to government expenditure due to a review of the methods used in the old system. Changes in the new system include:

  • the price indexes used to deflate social assistance benefits in kind were updated
  • the constant price measure for compensation of employees is now more in line with the current price measure. The new constant price measure deflates compensation of employees using an appropriate deflator. The previous method was to volume extrapolate using salaries and wages, which is only a subset of compensation of employees.

Revisions to government final consumption expenditure also occurred due to:

  • reconciliation to new annual current price measures as published in National Accounts (Income and Expenditure): Year ended March 2012
  • the introduction of FISIM (see section above) which increased final consumption expenditure.

Local government final consumption expenditure

Revisions to local government final consumption expenditure were due to:

  • the introduction of FISIM (see section above) which increased final consumption expenditure
  • a review of the local government final consumption expenditure system resulted in minor revisions to the constant price series before 1994.

Gross fixed capital formation

Revisions to gross fixed capital formation (GFKF) are due to reconciliation to new annual current price measures as published in National Accounts (Income and Expenditure): Year ended March 2012. The most significant revision to annual current price GFKF was in local government other construction, due to investment in road infrastructure.

Inventories

Inventories were revised due to reconciliation to new annual current price measures.

Exports and imports

Exports and imports were revised due to the implementation of FISIM (see section above). Exports of FISIM are slightly higher than imports. A review of the exports system resulted in minor revisions between June 2000 and March 2006.

Other revisions

In addition to the major changes listed above, other revisions were incorporated to the production measure of GDP this quarter, including:

  • Retail trade revised in the September 2011 quarter, due to revisions from the Retail Trade Survey
  • Revisions to professional, scientific, technical, administration, and support services; electricity, gas, water and waste services; and public administration and safety were due to updated indicators. All three of these updates revised the June 2012 quarter down.
  • Information media and telecommunications revised up in the June 2012 quarter, due to updated respondent data.
  • Revisions to transport, postal, and warehousing; and information media and telecommunications were a result of correcting errors in the chaining for these industries. These revisions were minor and did not affect total GDP.

The following table shows the previously published and revised quarterly movements for GDP and expenditure on GDP.  

Quarter Gross domestic product – percent change from previous quarter Expenditure on gross domestic product – percent change from previous quarter

 

 Previously published

Revised

Previously published 

Revised 

June 2007

0.5

0.8

1.7

 1.7

September 2007

0.6

0.7

0.9

 0.8

December 2007

0.4

0.2

0.2

 0.4

March 2008

-0.1

-0.4

-0.3

 -0.3

June 2008

-0.9

-1.0

-1.5

 -1.6

September 2008

0.0

-0.3

-0.3

 -0.3

December 2008

 -0.7

-0.6

-0.2

 -0.1

March 2009

-1.6

-1.1

-0.4

 -0.2

June 2009

-0.5

-0.4

0.8

 1.0

September 2009

0.3

0.6

0.2

 0.6

December 2009

0.9

1.6

0.7

0.4 

March 2010

0.5

0.1

0.5

 0.6

June 2010

0.6

0.8

0.2

 0.1

September 2010 

-0.1

-0.3

-1.0

 -1.3

December 2010 

0.0

-0.3

0.0

 0.2

March 2011 

0.6

0.7

0.4

 0.7

June 2011 

0.3

0.6

0.2

 0.7

September 2011

0.5

0.7

0.7

 1.1

December 2011

0.5

0.6

0.5

 0.8

March 2012     

 1.0

0.9

0.4

 0.4

June 2012

 0.6

 0.3

 0.3

 0.2

  • 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+