2  Definitions and measures of saving

Saving is generally defined as foregoing current consumption and providing funds directly or indirectly to capital markets for productive investment in financial, real or human capital (Boskin 1988). It can be measured in terms of “flows” as the difference between current income and expenditure.

Alternatively, saving can be measured as the change in the “stocks” of accumulated net wealth from one period to the next. Household net wealth, for example, can be thought of as the amount of money that households would be left with, if they sold off all their assets and paid off all their debts. Net wealth is thus the value of households’ assets (financial and real) less their debt, i.e. money owned on mortgages, vehicles, credit cards, student loans and other debt. Financial assets include assets with deposit-taking institutions, other fixed interest instruments, assets with life, super and managed funds, and directly held domestic and overseas equities. Real assets comprise housing and other tangible assets (although the Reserve Bank data does not yet include other tangible assets).

Saving rates in New Zealand are generally measured in terms of flows, derived from the System of National Accounts. However, there are important limitations in the measurement of saving rates as the residual between the flows of current income and expenditure.[2] Saving, measured as the difference between two large numbers (current income and current expenditure), is subject to wide margins of measurement error. This is because any errors in other series will be reflected in the saving estimate. For example, in 1998/99 household disposable income was about $58.1 billion and household consumption $60.5 billion. This implies a “dis-saving” of 4.1, as a percentage of disposable income. If income was under-estimated by, say two percent, and consumption was over-estimated by two percent, then the saving rate would have been zero instead of minus 4.1 percent.

Moreover, measured saving rates derived on this flow basis are biased downward. Expenditure on education, health care and durables is treated as consumption rather than investment, thereby understating saving. This downward bias is likely to increase over time as, for example, investment in education assumes greater importance. When allowances are made for educational investment and natural capital (e.g. mineral resources), New Zealand’s national saving rate in 1998, according to the World Bank, would have been almost six percentage points higher.[3]

An alternative measure of household saving can be derived from the net wealth data constructed by the Reserve Bank of New Zealand.[4] Household net wealth is also subject to measurement error and the derived saving rate is also likely to be biased downward. Household liabilities, which mainly consist of home mortgages, are almost certainly overstated.[5] This is because some home mortgages are in effect loans to small businesses (secured by residential mortgages). Some banks estimate that possibly between 10 and 20 percent of household loans secured on housing are for business purposes. Assets, on the other hand, are likely to be understated. Financial assets are understated as direct investment in non-financial substitutes, like forestry and farms, are not captured in the data (Thorp and Ung 2001). Moreover, the value of household investment in businesses not priced through the stock market is not captured. Real assets are also understated. Household real assets comprise house values alone and, for example, excludes consumer durables. In addition, net farm wealth, an important component of total net wealth in New Zealand, is excluded (Thorp and Ung 2001).