5.2  Household saving

International comparisons of saving across sectors are limited by data availability. One sector for which data are generally available is the household sector and much attention of the theoretical and empirical literature has focused on household saving.

Balance sheets or net wealth data are also available for Australia, Canada, France, Germany, Italy, Japan, the United Kingdom and the United States. Net wealth data for Australia are from the Reserve Bank of Australia, while the data for Canada, France, Germany, Italy, Japan, the United Kingdom and the United States were obtained from the OECD. The data should be more or less comparable across countries. However, one known difference is that the New Zealand data do not include other tangible assets, which are included in the other countries’ measures. Also, household assets in Australia include unincorporated enterprises. As a result, New Zealand household net wealth will be understated in absolute terms and relative to the other countries.

Figure 5.2 plots the flow and stock measures of household saving for the 9 countries that household net wealth data are available from 1986 to 2000. It shows that household saving measured by the difference between the flow of income and expenditure (bars) has been declining in all countries apart from France, at least over the 1990s.

A different picture emerges from the household saving rate implied by changes in household net wealth. As in New Zealand, the absolute differences between the flow and stock measures of saving are large. Countries that have less deregulated financial markets and where households are more credit constrained (France, Germany, Italy and Japan) appear to have a higher saving rate measured in terms of flows than in countries where financial liberalisation was largely completed by the mid to late 1980s (Australia, Canada, New Zealand, the United Kingdom and the United States).[19]

In the United States, the United Kingdom and Australia, the decline in the flow measure of saving was particularly sharp over the second half of the 1990s. In marked contrast, measured in terms of stocks, household saving increased over this period. In Canada and Germany, saving measured by the difference between the flow of income and expenditure also declined over the 1990s, while the stock measure remained virtually unchanged. The opposite occurred in Japan. The stock measure of saving fell and the flow measure remained basically unchanged.

The stock and flow measures of saving imply quite different pictures of household saving. Which of the two is the “right” measure to use depends, in part, on one’s beliefs about the determinants of households’ saving. There are three main theories of household saving behaviour: (i) the life cycle model of Modigliani and Brumberg (1954) and Ando and Modigliani (1963), (ii) the Barro (1974) model, and (iii) the Keynesian model.

Figure 5.2 - Stock and flow measures of household saving in selected OECD countries (as a percent of disposable income)

Source: OECD, Reserve Bank of Australia, Reserve Bank of New Zealand, Statistics New Zealand and authors' estimates.

Note: The flow measures of household saving for France, Italy and the United Kingdom and gross.

In the life cycle and Barro models, households choose how much to save and how much to work over their lifetimes. In order to equalise the discounted marginal utility of consumption from one period to the next, optimising households smooth consumption over time. Households save when income is high (during their middle years) and dis-save when income is low (during youth and retirement).[20]

The Keynesian model, in contrast to the life cycle and Barro models, is not based on an optimisation problem. Instead it simply assumes that households consume some fraction of their current income and some other fraction of their assets. While there is no formal justification for the Keynesian model, the informal argument rests on some combination of liquidity constraint and myopia.

Stock measures of saving take into account expected future earnings and are thus conceptually more closely related to the optimising-agent, life cycle and Barro models. Flow measures of saving, which only take into account current income and expenditure, are probably an accurate description of saving behaviour in the Keynesian model, where households simply consume some fraction of their current income and assets without regard for the future.

When considering whether households might be “adequately prepared” for retirement, the level of household net wealth (and associated stock measure of saving) is typically regarded as the more useful indicator than the current flow of household saving.