The Treasury

Global Navigation

Personal tools


Monthly Economic Indicators

Special Topic: Understanding the recent changes to the National Accounts

The gross domestic product (GDP) data released at the end of last year contained a number of revisions to the level and the growth rate of the New Zealand economy.  The level of nominal GDP was revised higher by about $4 billion for the year ended March 2014, due to a combination of better methods and more accurate information. Meanwhile recent growth in activity (real GDP) was revised downwards, suggesting that low inflation may not be as surprising as originally thought.

While revisions to the history of GDP data are normal (annual benchmarking occurs every year), these revisions were larger than normal. As well as the usual annual benchmarking, there was a rebase of the reference period and an update in the methods used to produce the national accounts. This note describes the effects of these revisions on the data.

Annual benchmarking improves measurement

These revisions to National Accounts result from balancing the production and expenditure estimates of GDP within a supply and use framework[1] using the most up-to-date information from the latest annual economic surveys. Also included were updated and new information from other data sources. Generally these improvements had small effects on the level and growth rates of the economy.

These changes, along with updating of the supply and use tables that underpin them, mean the data better reflect the evolving structure of the New Zealand economy.

Rebase affects the level of real GDP

Constant price series need a reference year on which to base the prices. In that year the value of nominal GDP (the value of activity in the economy) and real GDP (the volume of activity in the economy) are identical. Re-expressing the volume measures in a more recent base year will change the levels of real GDP (because we are using a higher nominal GDP as the reference) but will not contribute to any change in growth (Figure 1). The effect of prices is removed from nominal GDP to calculate real GDP.[2] When the reference year is updated (or rebased) then a different set of prices is used. In the latest data the reference period for prices has changed from the year ended March 1996 to the year ended March 2010, meaning for most activities prices have risen, raising the reported value of activity. [3]

Figure 1: GDP - Nominal revised up slightly, Real rebased
Figure 1: GDP - Nominal revised up slightly, Real rebased.
Source: Statistics NZ

A large part of this process is to update the weights applied at a more detailed level. These weights are used to combine detailed volume data (for example different mining products) that would have been held constant since 1996, until they were updated with the rebase. Overall, this did not have a marked effect on most of the economy as chain-linking is used to automatically update weights at higher levels.

However, prices of goods and services have changed at different rates. Changes in relative prices between base years mean the relative size of activities in the economy are different from the real shares using the earlier base year. For example, because the price of investments goods has declined relative to other goods, the domestic economy is larger as a proportion of the total than before. Meanwhile, exports as a share of GDP are smaller than previously (Figures 2 and 3). The value or amount of exports is not necessarily lower, but the proportion of total activity is not as great. Indeed, the benefit of an updated base year is that the relative sizes of the component series are closer to their actual contributions to economic growth.

Figure 2: Investment proportions better reflect relative prices
Figure 2: Investment proportions better reflect relative prices.
Source: Statistics NZ
Figure 3: Export share of GDP is smaller
Figure 3: Export share of GDP is smaller.
Source: Statistics NZ

Improved methodology increases coverage

Most significantly, Statistics NZ has upgraded the methods by which the national accounts are measured, moving from the previous international standards for measuring GDP and the balance of payments (SNA93 and BPM5) to the current ones (SNA08 and BPM6).  As the names suggest these changes are infrequent but are the internationally accepted norm for producing national accounts. Most developed countries are expected to have adopted the new standard by the end of the decade, with Australia, the UK, the US and Canada already using the new standards.[4]

Major changes fall into three categories: [5]

  • Classification of assets (widening the definition to include durable expenses such as research and development as assets and adding assets produced in-house);
  • Measurement of the financial sector (reflecting changes in the sector since 1993); and
  • Issues related to globalisation (reflecting the significant increase in cross-border transactions).

Further, improvements in measurement and the availability of more accurate data have been incorporated.  For example revisions to national saving are due to the availability of more accurate data for provisional estimates from the latest annual enterprise survey (2013) and updated tax data. As a consequence the saving rate is higher than previously estimated, especially among households (Figure 4).

Figure 4: Household saving revised up
Figure 4: Household saving revised up
Source: Statistics NZ

Nominal level revised up ...

The level of nominal GDP is around $4 billion or 1.9% higher in the year ended March 2014, so all else equal most measures will be lower as a proportion of GDP. For example, actual dollar values such as government debt remain the same but as a proportion of GDP falls from 26.2% to 25.6% at June 30 2014.  Similarly, New Zealand’s net international liabilities fall from 66.1% of GDP to 64.7% (also at June 30 2014), although better measurement of overseas assets has also contributed to this reduction in the debt position.

Household saving was revised up for the last five years. This was due to new data and some methodology changes by incorporating the new international standards. For example updated data largely resulted in increased distributions from unincorporated businesses to the household sector. The overall increase is a combination of both transfers from business saving to household saving and higher overall saving.

... and real growth revised down

While the value of activity has been revised up throughout history, recent growth in the volume of activity has been revised lower.

Lower growth in recent years may help explain lower inflation outcomes. Figure 5 illustrates the difference in quarterly growth rates in recent quarters. An implication of a slower growing economy is less inflationary pressure, consistent with CPI inflation undershooting expectations last year. On the other hand, if lower actual growth accounts for the lower inflation, then potential growth may not be as high as assumed previously, suggesting that inflation pressures will still emerge.

Figure 5: Recent growth revised lower
Figure 5: Recent growth revised lower.
Source: Statistics NZ

Measures of productivity and associated inputs of capital and labour will have changed as well. These changes are likely to be at the margin, but further analysis of the implications will be undertaken.

Overall, the Treasury’s view of the economy remains consistent with the outlook contained in the HYEFU (finalised before the new data were released). Data revisions are a regular occurrence, and while these recent changes were larger and more significant than usual, the underlying interpretation of developments in the economy is the same.

The New Zealand economy is growing at a pace slightly faster than potential with any spare capacity being absorbed (i.e. the output gap is close to zero). This has enabled growth to be solid without generating inflation pressure to date. However, future growth faster than trend will result in inflationary pressure developing (as a positive output gap grows). This pressure will be offset by developments in world commodity markets, especially lower oil prices.

The effect of lower oil, and other commodity, prices will be a feature of the Budget forecasts.


  • [1] This framework reconciles the flows of goods and services in the economy in a detailed analysis of the production and use of goods and services. It provides the basis for checking the consistency of the measures of the supply and use of goods and services, which have been estimated from quite different statistical sources
  • [2] In practice some volumes are adjusted for price changes to derive nominal GDP when the volume measure is easier to survey.
  • [3] Importantly, some prices have actually fallen in that time (such as mobile phones and computers), meaning for a given value of activity, there has been a greater volume of activity than previously published.
  • [4] Those countries that have adopted the new standards will likely appear to have larger economies and higher income per capita than those that have not. A comprehensive analysis of how the new measures affect cross-country comparisons and rankings will only be possible once all countries have upgraded their national accounts.
  • [5] For a full description, please see Statistics NZ’s Preview of 2014 national accounts improvements.
Page top