The Treasury

Global Navigation

Personal tools

Working Group

Saving New Zealand: Reducing Vulnerabilities and Barriers to Growth and Prosperity: Final Report to the Minister of Finance

8  Measurement issues

Measuring saving and net wealth is not a straight-forward exercise and is fraught with complex statistical issues. As discussed in Section 2.2, the SWG considers that the official national accounts summary statistics for national, government and private saving are reasonably robust, and are consistent with the related investment and external current account series. A problem – if there is a problem – lies with the subsequent split of private saving into its household and business components, even though the recent release of official saving sector statistics has reduced this source of uncertainty.[29]

There has been considerable debate about the level and trend of household saving. On the one hand, the official national accounts flow measure of household saving indicates, until recently, a declining and negative household saving rate; on the other, a number of recent studies using household survey data[30] find strongly rising levels of household wealth and in some cases have interpreted this as evidence of the exact opposite, namely a non-declining and positive saving rate. While definitional and data issues cloud a true comparison between the series, the major difference is due to asset revaluations (e.g., rising land and dwelling values). But even when attempts are made to remove the effect of revaluations, differences in estimates of household saving remain.

The SWG's prime focus has been national saving. While we necessarily discuss sector saving issues, we do so largely within a national saving context. As such, we do not attempt to address the conflicting household saving measures. We do, however, recognise this as an important issue that requires resolution. This is not some esoteric debate with minor consequences – on the contrary. Major public policy decisions with significant fiscal implications have been made, and will continue to be made, on the basis of evidence on household saving and retirement income adequacy.

This household saving measurement problem has been highlighted by the Retirement Commissioner and was also commented on in many of the submissions received. The recently released 2010 Review of Retirement Income Policy discusses household saving data needs and, in particular, notes:

“First, the current difficulty in reconciling available stock and flow saving data makes it challenging to draw firm conclusions about whether New Zealanders have a saving problem or in fact are saving adequately for their retirement. Greater reliability is essential for future policy making.

“Secondly, the outdated nature of current sources means it is not possible to assess recent trends and changes in saving behaviour. . . . this Review and other policy processes are drawing on outdated data. . . . Policy makers will require more timely data in future.” (Retirement Commission 2010)

The Retirement Commission report notes that Statistics NZ has recently reviewed its standard-of-living statistics, which include statistics on income, saving and wealth. Its review made a number of recommendations which are important for future understanding of the dynamics and trends in saving, wealth and retirement preparation.[31]The Retirement Commission report supported the Statistics NZ proposals and recommended that they be actioned in a timely way.[32]

While the Retirement Commission report has made this recommendation within a retirement-saving context, the statistics identified have wider application and the recommendation is fully endorsed by the SWG.

The need for improved saving and wealth data is not confined to the household sector. While the summary statistics on national and sector saving are now available, the absence of official sector financial accounts and balance sheets inhibits a richer analysis of the types of economic problems that have been discussed in this report. A better understanding of the connections between the levels and types of investment, their internal and external (inter-sectoral) funding sources and the resulting balance sheet positions is needed.

More generally, the work of the SWG has been made more difficult by a lack of timely economic statistics on sector saving and net wealth. While progress has been made with the release of official institutional sector accounts, these tell only part of the story.[33]As noted above, there is a need for institutional financial accounts and balance sheets to better understand the nature of the economic issues under review. We note that such statistics are now regarded as a standard statistical output in other OECD countries and, in this regard, New Zealand stands out as an outlier. This situation would appear to be the inevitable consequence of successive governments' under-investment in providing and maintaining an adequate range of quality economic statistics. The SWG strongly recommends that these deficiencies be addressed.

Specific recommendations:

1  Macroeconomic statistics

That Statistics NZ:

  1. Improve the timeliness and periodicity of the recently released institutional sector accounts. The goal should be to release the annual accounts within six months of the end of the financial year.
  2. Develop the full sequence of accounts for each institutional sector, as set out in the international System of National Accounts manual. This involves developing financial accounts, balance sheets and reconciliation accounts, in addition to the existing income and outlay and capital accounts currently produced. The production of these accounts with a quarterly frequency is strongly recommended.[34]
  3. Continue to research macro and micro household saving measures to enable a fuller reconciliation between them.

2  Household sector statistics

Household sector. The SWG endorses Recommendation 5.5 of the 2010 Review of Retirement Income Policy and supports the relevant recommendations in the Statistics NZ Review of Standard of Living Statistics, namely:

  1. Work be undertaken to better integrate the macro and micro measures of household saving.
  2. Cross-sectional statistics on household net worth be regularly produced.
  3. A new longitudinal survey be investigated now that the Survey of Family, Income and Employment (SoFIE) has ended.


  • [29]Statistics New Zealand (2010a).
  • [30]Refer, for example, to: Claus and Scobie (2002), Gibson and Scobie (2001), Le, Scobie et al (2009), Scobie and Henderson (2009) and Le, Gibson et al (2010).
  • [31]The relevant Statistics NZ recommendations are: (i) Statistics NZ to undertake a comparison of micro and macro measures of income and wealth with a view to better integrating these and a goal to work towards a full set of national accounts; (ii) Statistics NZ to confirm information need in relation to household net worth and investigate options for producing regular statistics; and (iii) Statistics NZ to work with users to prioritise needs for longitudinal data and investigate options for meeting these needs, including the potential role for future longitudinal surveys. Recommendations (ii) and (iii) are in response to the need for an update to the 2001 Household Saving Survey which provided cross-sectional data on household assets and liabilities, and for a replacement for the longitudinal survey SoFIE which has now ceased.
  • [32]Recommendation 5.5. “That the recommendations relating to saving and wealth statistics from Statistics NZ's review of economic living standards be actioned in a timely way and in conjunction with key data users.” (Retirement Commission 2010, p107)
  • [33]The institutional sector accounts include only the income and outlay account and the capital account.
  • [34]The recent GFC has led to an international review of the adequacy of official statistics to provide the relevant and reliable information needed for timely economic analyses. What is emerging is a recommended list of key early warning indicators that include, inter alia, quarterly statistics on sector incomes, profits, saving and debt (levels and changes). The full sequence of quarterly institutional sector accounts proposed above would provide this data.
Page top