The Treasury

Global Navigation

Personal tools


Monthly Economic Indicators

Special Topic 1: Revised New Zealand GDP Data

Statistics New Zealand released revised Gross Domestic Product (GDP) data on 15 May 2012 that made reasonably significant changes to how we view the economy in recent years.  These changes were made in the week before BEFU 2012 so were not included in the economic and fiscal forecasts.  This note outlines some implications of the revised data.

A number of changes were made to GDP...

The revisions affected the two measures of real quarterly GDP available in New Zealand: production (ie, value-added output by industry) and expenditure (ie, consumption, investment, exports and imports).  Production is the headline measure in New Zealand.  The changes incorporated a new industry classification (the Australian and New Zealand Industrial Classification or ANZSIC 2006) and methodology improvements, plus updated annual benchmarks and chaining weights.  The largest impact was from the new annual benchmarks. 

The main feature of the revisions is a reduced gap between the two measures of real GDP.  These measures should give the same result in theory, but prior to the revisions the expenditure measure was 4.5% higher than production in the December 2011 quarter (Figure 6).  With the revisions, the gap is now 0.9% as expenditure was revised down 2.9% and production moved up 0.6%. 

Figure 6 – Gap between expenditure and production
Figure 6 – Gap between  expenditure and production.
Sources:  Statistics NZ, Australian Bureau of Statistics

...but still a gradual recovery from recession

The new data do not change the gradual pace of recovery following a moderate contraction in 2008/09, but in confirming this they were rather disappointing (Figure 7).  There was some hope that production GDP would be revised much higher than it actually was.  The size of the recovery is even more gradual than previously at 3.4% since September 2009 (rather than 3.6% since June 2009).  Nevertheless, real GDP has not contracted over the past 10 quarters, a record which only eight other OECD nations can claim.  The December 2010 quarter is now barely a rise, up just 0.003%, so it is prudent to say 9 of the last 10 quarters have shown growth.

Figure 7 – Real GDP revisions
Figure 7 – Real GDP  revisions.
Source:  Statistics NZ

The size of the contraction in 2008/09 is now 3.3% from a peak level of activity in the December 2007 quarter (the pre-revision fall was 3.7%).  However, the nature of this contraction is not as straightforward as previously – five consecutive quarters of declining real GDP from March 2008 to March 2009 now become two consecutive declines in the first half of 2008 followed by a steady quarter (0.01%) in the September 2008 quarter and then three further quarters of decline.  This gives us two distinct recessions on the so-called "technical" definition of two consecutive quarters of contraction – the first a domestic recession induced by drought, the collapse of finance companies and tight monetary conditions, and the second an external recession after the global financial crisis in late 2008.

The size of the recovery (3.4% or $1.143 billion) in quarterly real GDP is almost identical to the size of the contraction (3.3% or $1.154 billion), hence the level is just $11 million below its all-time high.  Also, annual average growth in real GDP was revised up from 1.2% to 1.5% for 2010 and revised down by an equal amount from 1.4% to 1.1% for 2011.  With growth in 2011 now weaker than in 2010, the economy had less momentum heading into 2012 than previously thought.

Significant changes at detailed level...

The introduction of ANZSIC 2006 was partly to improve the measurement of services in the economy.  Interestingly, the size of the service sector was revised down (from 71% of GDP to just under 70%) as some industries moved into the goods-producing sector, which is now quite a bit larger.  Agriculture was revised down, partly because the 2008 drought was better accounted for (although there is now no recovery from this drought).  There were only minor changes to the tradable and non-tradable split presented in the last Monthly Economic Indicators (Figure 8).

Figure 8 – New tradable and non-tradable output
Figure 8 – New tradable  and non-tradable output.
Source:  Statistics NZ, the Treasury

Around half of the revision to real expenditure GDP was accounted for by a reduction in private consumption (2.3% lower in the December 2011 quarter), partly caused by a 1.8% increase in the deflator (ie, consumer prices).  Some of this appears to reflect a fuller flow-through from the rise in GST in 2010 to be more consistent with what we expected at the time, while the increase in 2011 seems out of line with the Consumers Price Index.  Nominal expenditure GDP was revised down by 1.5%, of which 0.8% percentage points reflected a revision to public consumption.  This revision seems inconsistent with growth in central government spending in recent years.  The rest of the downward revision in nominal expenditure GDP came from weaker private consumption and investment. 

...could have implications for potential output

Potential output and output gap measures based on production GDP may not be greatly affected given the small 0.6% upward revision to this series.  For example, average labour productivity growth in the 15 years to March 2011 is now 1.2% per annum, up slightly from 1.1% previously.  Of concern is the slightly weaker recovery since the recession ended and what this implies for future growth.  However, it is interesting to note the immediate recovery in 2009/2010 is now stronger but growth rates since the 4 September 2010 earthquake have been weaker.  This could give more weight to our initial estimate of a 1.5 percentage point hit to growth in 2011 from the earthquakes, which would suggest growth has a lot of room to bounce back.  Measures of potential output built up from expenditure GDP will be affected negatively by the revisions.  This includes the private sector estimate of potential output used in the New Zealand Treasury Model (NZTM), although the impact will be lessened by the government sector taking some of the hit.

Key lessons and implications

No GDP number is infallible: they are all subject to change.  Revisions are an "occupational hazard" for economic forecasters, but the latest changes are large enough to call into question quarterly estimates of GDP.  The December 2010 quarter has been revised from 0.2% on first release to 0.5% to 0.6% back to 0.3% and now 0.0% (our original pick in the preliminary BEFU 2011 forecasts).  The September 2008 quarter was revised from -0.5% to 0.0%, our original pick at the time. Despite the revisions, there are still differences between the two measures of GDP.  Did real GDP fall 0.6% in the June 2009 quarter as the production measure suggests, or did it rise 0.8% as the expenditure measure shows?  Although the levels gap is much smaller than previously, the correlation of quarterly growth rates between the two GDP measures has deteriorated over the period 2007-2011 since the revisions.

The revised data have little direct impact on BEFU  forecasts, but there could be some effect depending on how the changes are interpreted.  Even if we keep the same quarterly growth going forward, annual average growth for 2012 would be lowered from 2.2% to around 2% (2013 would remain at 3.5%).  If the revisions do suggest either less momentum heading into 2012 or a lower level of potential than we previously thought, then these growth rates would be lower still.  Fiscal numbers presented as shares of GDP will change slightly from those presented in the 2012 BEFU – a surplus will stay a surplus but would appear larger if the downward revisions to nominal GDP continue to flow through, while debt levels as a share of GDP would be slightly higher.  Tax revenue forecasts may not be impacted much or at all if nominal economic growth forecasts remain unchanged.  We will continue to analyse the revised GDP data and its implications for our forecasts and understanding of the economy.
Page top