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+
Household Economic Survey population rebase: year ended June 2007 to 2015

We have revised estimates for the Household Economic Survey (HES), using information from the 2013 Census. We outline the changes between the originally published and revised estimates, and the revision of regional benchmarks to help improve the accuracy of our estimates. 

Population rebase

The HES is a sample survey. From this, we use statistical weights to calculate income and expenditure estimates for the total New Zealand population. Following each census we revise the weights, to make use of the latest population counts. This process is called a population rebase. In the latest rebase, we revised the statistical weights from the year ended June 2007 to the current release (year ended June 2015).

Revision of regional benchmarks

In HES, regional benchmarks enable us to create better estimates of income and expenditure by region. This year, we have reduced the number of regions benchmarked in HES from 16 regional council areas to 5 broad regions. This is because the HES sample size is too small to create meaningful estimates for all regions, and using benchmarks at a more-aggregated regional level will enable better accuracy.

Updated HES estimates

The Excel tables in the Household Economic Survey (Income): Year ended June 2015 information release are updated with the revised estimates. We do not recommend comparing the results in the 2015 release with the previous HES release in 2014.

See NZ.Stat for the revised time series for the expenditure tables.

Revisions from the rebase and regional benchmark change

At the national level, the 2013-base estimated resident population was consistently less (within 0.7 percent) than the 2006-base estimated resident population.

Income and expenditure estimates are affected by the population rebase and the regional benchmark change.

Population change

The population rebase resulted in:

  • a downward revision in the working-age population (particularly for males)
  • a downward revision for people aged under 35 years
  • an upward revision for people aged 35 to 50 years


The rebase and benchmark changes resulted in minor revisions to the income estimates. For our key income estimates in all years (2007–14) we found:

  • household income from all regular sources was revised by less than 1.1 percent
  • personal income from all regular sources was revised by less than 1.2 percent

Table 1 summarises changes in the key measures between originally published figures and revised estimates for the year ended June 2014.

Table 1

Changes in average annual household income, by income source
Year ended June 2014

 Average annual household income from:

 Original ($)  Revised ($)  Percentage change

 Wages and salaries

 88,357  88,277  -0.1


 50,191  49,515  -1.3 


 10,700  10,555   -1.4

 Private superannuation

 24,178  23,249  -3.9 

 New Zealand Superannuation and war pensions

 23,654  23,218  -1.8 (1)

 Other government benefits

 11,601  12,596   8.6 (1)

 Other sources

 17,903  18,257   2.0

 Total regular and recurring income

 88,579  88,835   0.3

1. These changes were not solely the result of the population rebase and change to regional benchmarks.
    See revisions to government transfer income for more information

Estimates based on small sample sizes were more affected by the revisions (eg irregular income, and income from private superannuation). However, of all estimates of personal and household income (for any year, from any income source), we revised only a single estimate significantly (ie change was larger than the sample error).

This means the rebase and new benchmarks have not changed our understanding of income within New Zealand.

The estimate that changed significantly was the 2014 estimate of median annual income from wages and salaries for all people (this includes people without jobs). The original estimate was $18,480; after the rebase and new benchmarks, this was revised downwards by 4.8 percent, to $17,598. The move is largely explained by the downward revision of the male working-age population.


As with income, the population rebase led to minor revisions in household expenditure estimates for each year from 2007 to 2014. Any variation between the published and rebased expenditure figures is only nominal.

For our key expenditure estimates in all years (2007–14) we found:

  • total average weekly expenditure on housing costs revised by less than 1 percent
  • net weekly household expenditure (measured in 2007, 2010, 2013) revised by less than 0.5 percent
  • at a regional level, overall household expenditure revised by less than 2 percent.

Table 2 shows the effect of the rebase on the average weekly expenditure on housing costs (for households reporting that expenditure) between 2007 and 2014. This is largely in line with the changes in the overall population.

Table 2

 Average weekly expenditure on housing costs
 Year ended June 2007–14


 Original ($)  Revised ($)  Percentage change


 209.10  209.80   0.3


 235.30  235.10  -0.1


 238.30  238.20   0.0


 236.00  235.00  -0.4


 254.50  254.00  -0.2


 255.90  254.90  -0.4


 259.60  257.10  -1.0


 284.50  284.00  -0.2


  • 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+