Housing: An Analysis of Ownership and Investment Based on the Household Savings Survey

In 2001, Statistics New Zealand conducted a major survey of the assets and liabilities of New Zealanders called the Household Savings Survey (HSS). This paper presents the results of an analysis of ownership and investment in housing based on the results of that survey. International comparisons suggest that the rates of home ownership, investment in property and housing debt levels in New Zealand are broadly comparable with those in Australia and the United States and with a wider set of countries. An exception is that younger age groups in New Zealand hold more investment property than their counterparts in the USA and Australia. In New Zealand almost one in ten couples owned rental property in 2001, while one in five owned some form of investment property. We examine the factors that govern tenure choice and gearing. Of note is the fact that 44% of couples and 56% of individual home owners have debt free residential properties. Households' balance sheets reflect the importance of housing for both assets and liabilities. We complement the analysis of the cross-sectional unit record data from the HSS with an analysis of housing taken from the households' aggregate balance sheets from 1978 to 2004 from the Reserve Bank of New Zealand. We use these data to form a measure of household saving based on the stock of net equity. We then adjust this measure of savings for changes in house prices, and find that this adjustment explains almost two thirds of the difference between the stock and flow measure of household savings, the latter taken from the Household Income and Outlay Accounts. Furthermore we find that from 1980 to 2005 the annual average rate of household saving based on these estimates from household balance sheets was 12.4% of personal disposable income, after removing the effect of changes in house price. Arguably this is a preferable measure of household saving to the widely cited negative rates of household saving based on national income accounts. We further use the balance sheet data to estimate the extent to which households have apparently withdrawn equity from their housing assets for investment in other forms or consumption. We find that on average a rise of one dollar in housing net equity is associated with 10 cents of apparent equity withdrawal.