RDP 2022-06: Do Australian Households Borrow to Keep up with the Joneses? 1. Introduction

There has been increasing academic and policy interest in the effect of inequality on both microeconomic and macroeconomic outcomes. One aspect that has received attention in the literature is the potential for local inequality to affect indebtedness and credit allocation. A number of papers have found that a high level of local inequality leads to increased levels of indebtedness, as households appear to try to ‘keep up’ with the income and consumption of their wealthier neighbours (e.g. Georgarakos, Haliassos and Pasini 2014; Berlemann and Salland 2016). This has potential important implications for both financial and macrofinancial stability if this ‘keeping up with the Joneses’ leads households to become overextended, and more sensitive to negative shocks. Other papers have argued that local inequality can limit the allocation of credit to lower income households in a region if banks assess them relative to their local peers and have geographic caps on lending (Coibion et al 2020; Loschiavo 2021). Such screening effects can lead to an inefficient allocation of credit and entrench inequality.

In this paper, I examine the effect of local income inequality on debt using Australia as a case study. Australia shares many similar economic and financial conditions with the United States and Europe, the focus of existing literature, but with much higher levels of household debt. Most analysis on income inequality in Australia has focused around the national level, which has remained broadly flat in the past decade (Productivity Commission 2019). The long streak of uninterrupted economic growth in Australia has brought significant improvement in income for the average Australian household in every income decile, in contrast with the United States where income inequality increases at a similar rate as Australia but the distribution is much more uneven (Productivity Commission 2019). However, aggregate inequality masks variation at the local level, which may serve as a key reference group for most individuals when making decisions. The significant spatial and temporal variation at the local level makes Australia an interesting case study.

I find evidence of ‘keeping up with the Joneses’-type dynamics in Australia. Specifically, there is a significant positive association between local income inequality and household investment debt. The accumulation of investment debt is mainly driven by middle-aged and middle-income mortgage holders without liquidity and borrowing constraints and who are willing to take financial risk. Car debt also increases moderately with a rise in local inequality, driven by middle-aged outright home owners who self-perceive as financially prosperous. The accumulation of car debt lends evidence to the traditional ‘conspicuous consumption’ channel, with households trying to close consumption gaps in a conspicuous manner. On the other hand, the accumulation of investment debt suggests an additional channel for the ‘keeping up with the Joneses’ hypothesis – households borrow to invest, in the hope of keeping up with a rising local income gap. Such a channel could be particularly concerning if the households that take on the additional risky investments were financially fragile. However, in Australia it appears that they are taken on by households that are financially comfortable, lessening such concerns.

The paper directly relates to the literature examining the relationship between household indebtedness and income inequality. One strand of this literature identifies positive peer effects on borrowing, suggesting a ‘demand’ channel. This relates to the well-known ‘keeping up with Joneses’ phenomenon often cited as an explanation for households' excess of labour supply and overspending in the United States (Stiglitz 2012). According to this hypothesis, which goes back to Veblenʼs theory of conspicuous consumption, an increase in income inequality prompts lower-income groups to spend or borrow more in trying to imitate the consumption patterns of higher-income groups. Georgarakos et al (2014) show that Dutch households who consider themselves poorer than their reference group borrow more. Similarly, Berlemann and Salland (2016) find positive effects of the local average income on debt market participation using German banking data. Another strand in the literature supports the ‘supply’ channel in which creditworthiness is private information and local income inequality serves as a ‘signal’ to screen borrowers. This leads to credit being more readily accessible to high-income households in more unequal areas, as found by Coibion et al (2020) using US credit data. Likewise, Loschiavo (2021) shows that the probability of being indebted is higher for high-income Italian households in high-inequality areas compared with equivalent households in lower-inequality areas, suggesting that as local inequality increases banks are more likely to use a household's position in the income distribution as a measure of creditworthiness.

This paper contributes to the literature in several ways. First, I explore the effects of changes in local income inequality on not only total household debt but also its composition, including mortgage debt and different types of non-mortgage debt. Second, I examine the heterogeneity of the effects across different cohorts based on demographics and financial capacity, which, to my best knowledge, has not been done in the literature before. The very rich micro data on Australian households from the Household, Income and Labour Dynamics in Australia (HILDA) Survey, which includes questions on the details of household balance sheets, allows for this granularity. By exploring the composition of debt, as well as the heterogeneous effects across household types, I shed substantial additional light on the mechanisms and macroeconomic implications of the relationship between local inequality and household borrowing.