RDP 2005-10: Housing and the Household Wealth Portfolio: The Role of Location 1. Introduction

For most Australian households the most valuable asset is their own home, with the average owner-occupier holding around 60 per cent of total assets in housing. This portfolio mix varies considerably across all households, and more specifically, it varies between urban and rural locations.

Based on wealth data for owner-occupiers in the 2002 Household, Income and Labour Dynamics in Australia (HILDA) Survey, Figure 1 shows that, on average, households in urban areas hold a higher share of their assets in their own home than do rural households (see the Appendix for a detailed description of the data and sources).

Figure 1: Average Share of Total Assets Held in Own Home

The question then is why might urban households be willing to allocate a higher share of their assets to housing? Presumably there must be some factor, or combination of factors, that can explain such behaviour, since by itself holding a higher share of assets in the own home is a costly exercise. This follows from the fact that a portfolio which is heavily concentrated in the own home would limit the benefits that might otherwise come from a more diversified portfolio of assets, such as lower risk and possibly higher income.

There are three possible benefits that might offset this cost. First, households may be willing to pay a higher price for a home in an urban area if the house itself is assessed as being of inherently higher quality. However, as we argue later, there is evidence to suggest that this is not the case, on average.

A second possibility is that higher dwelling prices in cities could reflect the well-known ‘urban wage premium’ (Glaeser 1999).[1] That is, controlling for household characteristics (their earning and saving capacity in particular), a household can expect, on average, to earn a higher income and accumulate more wealth in a more urban setting. Faced with a choice between an urban and rural setting, households should be willing to pay more for their home in order to secure access to the higher incomes which are generally available in cities.[2] In principle, this could lead households to hold a higher share of assets in their own home in more urban settings. However, this need not be the case and, as we discuss later, this effect could even work in the opposite direction.

A third possibility is that there is an ‘urban premium’ built into house prices reflecting the other (non-pecuniary) benefits that cities can provide, which could explain why households are willing to hold a higher share of assets in their own home. These benefits could include greater access to infrastructure and services (such as health and education), as well as the opportunities to interact with a larger pool of people. At the same time, however, there could be costs associated with greater urbanisation related to congestion, including increased travel times, greater pollution and more crowded public facilities.

In short, the fact that urban households are willing to devote a larger share of their total assets to housing suggests that in net terms they value the urban setting. How much they value urbanisation, and whether this also reflects non-pecuniary benefits of cities, is the focus of this paper.

Using a recent cross-section of wealth data from the HILDA Survey, we examine how households living in different locations allocate their total asset holdings. Looking at owner-occupiers only, we test whether urbanisation is a significant predictor of the share of total assets held in property, once we control for income and other factors that can affect asset allocation. For owner-occupiers, this effect is found to be significant and positive, with a 100 person per square kilometre increase in urbanisation increasing the share of assets held in the home by 0.4 percentage points, on average. We also find that this effect is not linear but declines at higher levels of urbanisation. Hence, for example, for an average household moving from Cairns to Brisbane city – an increase in urbanisation of around 2,000 persons per square kilometre – the increase in their housing share of total assets is estimated to be only 5.6 percentage points.

We extend our analysis to account for two aspects of home ownership: the consumption of housing services and the investment decision. To shed some light on this distinction, and how location affects this, we analyse households that own their home and at least one other property. In this case, the effect of an increase in urbanisation is found to be (close to) linear for their own home. In contrast, the investment component of housing shows that there is a limit, albeit quite high, to the share of total assets that multiple property owners are willing to allocate to total property assets.

The remainder of this paper is structured as follows. In Section 2, we briefly examine previous studies related to the subject of our paper. Section 3 examines the role of housing in the household portfolio for our dataset. Section 4 discusses our choice of urbanisation measure and presents the estimation results. Section 5 concludes.


Closely related to this, the greater employment and business opportunities in more urban areas will tend to reduce the variability of wage and business incomes, thereby offsetting some of the risk associated with a less diversified portfolio of assets. [1]

While this discussion pre-supposes that households make the same tenure decisions independent of location, such an assumption seems reasonable given that home ownership rates are equally high across rural and urban locations, at around 70 per cent. [2]