Financial Deregulation and Household Debt: The Australian Experience

Abstract

Australian households have become increasingly indebted over the past 20 years. Figure 1 shows various measures of household indebtedness rising almost continuously since the late 1980s. Household debt as a proportion of disposable income rose from around 45 per cent in 1988 to just below 160 per cent in 2008. Most of this debt has financed investment in housing (housing currently accounts for 87 per cent of household debt), predominantly owner-occupied housing but, more recently, also housing for rental purposes. Borrowing to invest in rental housing has been the greatest contributor to the growth of household debt in recent years. Rising Household Debt in Australia: Key Ratios (%) Note: LHS denotes left hand side; RHS denotes right hand side.Source: RBA (2009). Although house prices increased over the same period, along with asset prices generally, they did not keep pace with rising levels of household debt, raising the debt-to-assets ratio from 8 per cent to slightly less than 20 per cent. Interest payments as a proportion of disposable income were relatively constant over most of the period, as falling interest rates largely offset higher debt levels. But, this ratio too began to rise as interest rates picked up over the first half of the decade since 2000. Australia is not unique, as most Organisation for Economic Co-operation and Development (OECD) countries have experienced similar rises in household indebtedness in recent years, although the levels in Australia are now higher than those of most developed countries. This article examines the link between financial deregulation and higher levels of household debt, as well as the implications for inter-temporal choice and macroeconomic stability. Financial deregulation in Australia commenced in 1980 with the removal of interest rate ceilings on bank deposits and all bank loans, apart from mortgage loans for owner-occupied housing.1 Prior to deregulation, financial intermediaries were subject to a range of controls on interest rates, reserve requirements, liquidity ratios, portfolio constraints and restrictions on market entry, particularly by foreign lenders. At the macroeconomic level, the exchange rate for the Australian dollar was fixed and foreign exchange transactions were subject to quantitative controls. Over time, these controls increasingly distorted resource allocation within the financial system and undermined the effectiveness of macroeconomic policy. The case for financial deregulation strengthened (Harper 1986). Financial regulation affected Australia's financial institutions and markets in a variety of ways, including: restrictions on banks weakened their competitive position relative to non-banks: by the early 1980s, the market share of banks had fallen to 40 per cent, compared with 70 per cent in the early 1950s (Battellino 2007); controls on entry stifled competition and, hence, efficiency and innovation, resulting in wide interest rate margins by international standards; new intermediaries and sophisticated financial instruments developed to circumvent controls; interest rate ceilings often led to credit rationing at a cost to both efficiency and equity; and inadequate interest rate flexibility and the falling market share of banks undermined the effectiveness of monetary policy. For the household sector, in particular, the effect of financial regulation was to restrict Financial market reform began in earnest with the establishment of the Campbell Committee in 1979. The Campbell Report recommended a series of major macroeconomic reforms, such as floating the exchange rate and introducing a tender system for the sale of government securities. The Committee also recommended a number of changes that directly affected the household sector. Chief among these were the removal of barriers to entry, including the de facto ban on the entry of new banks, and the removal of direct controls on interest rates and asset allocations (Harper 1997). As the Wallis Report later observed, the majority of the recommendations contained in the Campbell Report ultimately were implemented. Following the Campbell Committee, the Martin Review Group was commissioned in May 1984 to ‘… report on possible changes which might be made in the framework of official regulation and control of the financial system’ (Martin et al. 1984, p. 1). The Group broadly endorsed the view of the Campbell Committee, albeit with greater emphasis on the practical issues of implementation. The entry of new banks and a relaxation of ownership conditions were recommended, along with the removal of a wide range of interest rate controls, including those on housing loans. In contrast to the earlier reviews and reflecting the widespread financial reform that had occurred over the preceding decade, the Wallis Committee (1996–97) did not focus on deregulation. Rather, its brief was to reconfigure financial regulation and regulatory institutions so as to facilitate more appropriate commercial responses to changing financial conditions while preserving financial system safety and stability (Harper 1997). The Wallis Report's recommendations had less direct impact on the household sector, yet contributed to a more flexible, efficient and stable financial system. This is evident in the soundness of Australia's financial system before and during the current global financial crisis. the supply of credit and the choice and diversity of financial products (Harper 1991). Credit-worthy borrowers were left with unsatisfied demand for credit and the level of household debt was low by international standards. The ability of households to smooth their consumption profile over time was constrained by what amounted to a regime of official credit rationing. In response to the pressures and inefficiencies of financial regulation, successive Australian governments, through a series of public inquiries (see Box 1), identified key areas for reform. The resulting deregulation of financial markets in Australia was broadly similar to that adopted throughout the OECD area, although the timing, pace and scope of reform varied across countries. Financial deregulation affected households in several ways. First, deregulation opened up the lending market, which was no longer constrained by the price and quantity controls previously in place. Second, interest rates fell, spurred on by greater competition as new (including foreign) banks entered the market and existing lenders crossed into each other's territory. The Reserve Bank of Australia (RBA) estimated that financial deregulation was responsible for a fall in net interest rate margins of around 2 percentage points over the two decades to 2003 (RBA 2003). Greater competition also manifested in a broader range of product offerings to consumers, meaning that mortgages were not just more accessible, but that product choice was more diverse.2 Credit cards also became more widely available. Much of the increase in household debt following deregulation is simply the result of households accommodating themselves to the absence of artificial credit rationing. In earlier times, households were denied access to sufficient credit to adopt their preferred allocation of consumption through time. Moreover, households have gained significantly from enhanced freedom to leverage their asset purchases in the form of higher levels of household wealth (see Figure 2 below).3 To conclude that households have had too much of a good thing—that is, they are over-borrowed—there must be a reason to think that their inter-temporal choices are systematically distorted or that the aggregate outcomes of individually rational decisions are collectively sub-optimal. In other words, there must be either good microeconomic or macroeconomic reasons to believe that, free of artificial credit rationing, households will borrow too much for their own good or for the good of the economy. Movements in Household Net Worth and House Prices (Indexed: 1989 = 100) Sources: ABS (2008b); ABS (2009); Department of Parliamentary Services (2006). Notwithstanding the significant growth of household indebtedness, the mortgage default rates in Australia have remained consistently low despite large increases in debt-servicing requirements. Similarly, there has been no obvious increase in the rates of household bankruptcy over time (see Figure 3 below). Increasing leverage exposes households to greater risk, but also to the expectation of earning higher returns on their investments. Prior to deregulation, Australian households would appear to have been constrained to levels of risk that were lower than their preferred exposure, simultaneously lowering their expectations of return on invested funds. In any case, which market distortions might induce households to borrow more than they would otherwise? Non-Performing Loans by Country (%) Source: Unpublished RBA data. One possibility is the preferential tax treatment of owner-occupied housing compared with other investment classes, including rental housing. Imputed rental income from owner-occupied housing is not taxed; nor are capital gains. However, mortgage interest payments on owner-occupied housing are not deductible, unlike in the United States, for example. On balance, such treatment for income tax purposes favours the investment in owner-occupied housing over other asset classes and favours equity over debt as a means of financing such investments. Therefore, one might expect Australians to live in larger houses than otherwise, but also to see them paying down their mortgages as quickly as possible. However, there is no reason to expect households to be more indebted than they might be otherwise. They would still borrow as much as possible, given their ability to service the debt and their willingness to risk unemployment or some other impediment to their ability to repay. At worst, the tax distortion induces Australians to over-invest in housing rather than over-borrow to finance such investment. Of greater potential concern are the macroeconomic consequences of households collectively increasing their levels of debt. Increasing leverage inevitably exposes borrowers to higher risk; indeed, this is the basis of any expectation of a higher return on investment. With the household sector, in aggregate, carrying more debt, there is a greater exposure to unexpected reductions in future income growth. When debt levels are lower, households that experience reduced expectations of future income growth can adjust to their diminished prospects by reducing both consumption and saving. When debt levels are higher, a greater share of future income has been pre-committed to saving and debt-servicing. Therefore, an unexpected decline in future income will exert a concomitantly higher negative impact on household consumption. In other words, higher levels of household debt potentially increase the volatility of household consumption in the face of unexpected volatility of income. Given that aggregate private consumption is around 55 per cent of the national income, more volatile consumption potentially induces a more volatile national income. Unanticipated increases in interest rates have a similar effect. Higher interest rates exert a stronger contractionary effect on consumption when a higher proportion of income is committed to servicing debt. The effectiveness of monetary policy is sharpened accordingly, with smaller increases in interest rates required to achieve a given reduction in aggregate demand. So, although aggregate demand is potentially more volatile at higher levels of household debt (as the impact of demand shocks is transmitted to a greater extent to consumption, which feeds back onto aggregate demand), the potency of aggregate demand management through monetary policy is similarly enhanced. Although the disease is more virulent, the cure is more efficacious. If the shock to incomes is sufficiently severe, higher levels of household debt can transmit this instability to asset prices, as households are forced to reduce not just their consumption but also to liquidate assets. The more indebted households are, the closer they come to this threshold. Again, higher levels of household debt increase the potential for negative income shocks to become self-reinforcing, both through transmitted shocks to consumption and to wealth via asset prices. Estimates of the wealth effect on consumption vary but a loose rule of thumb is that a doubling of wealth leads to an increase in consumption in the order of 5 per cent.4 Reductions in wealth may give rise to larger cuts in consumption, given the threshold effect of bankruptcy. These effects are muted if the majority of household debt is incurred by wealthier and/or higher-income households that enjoy wider margins to absorb shocks to income and wealth. There is some evidence to suggest that this might be the case in Australia. Events of the past 18 months have played out the risks to which higher levels of household debt expose the Australian economy. Wealth has fallen dramatically as asset prices have tumbled (the aggregate value of household assets fell by 13 per cent in the year to December 2008) and unemployment has risen, leading to a sharp loss of income for some households. To date, however, there are no signs that the economy has been destabilised seriously by the indebtedness of the household sector. An important buffer has been the dramatic drop in interest rates as monetary policy has eased, which has significantly reduced households’ debt-servicing commitments. The RBA estimated that, over the 6 months to March 2009, household interest payments fell by the equivalent of 5 per cent of disposable income (Edey 2009). There are indications that households are responding differently to current conditions than they did under less-indebted circumstances. In particular, household consumption is falling more quickly as households seek to reduce their debt levels in the face of falling income and asset prices. Given the additional time it will take to repair household balance sheets, the recovery of household consumption is also likely to be delayed. The most recent data from the Australian Bureau of Statistics on the household savings rate show a sharp increase, in marked contrast to the experience in previous recessions. Compared with the 1982–83 recession, when the savings rate fell by around 2 percentage points, and the 1990–91 recession, when the savings rate fell by 40 per cent, the early stages of the current recession have seen an increase in the savings rate, from 1.2 per cent in the March quarter to 8.5 per cent in the December quarter of 2008. Rather than smoothing consumption during the downturn, households appear to be deleveraging their balance sheets to reduce exposure to further unanticipated shocks to income and wealth (Figure 4). Savings as a Share of Household Disposable Income (%) Source: ABS (2008a). As consumption falls further and more rapidly in the face of higher household debt levels, the gross domestic product will follow suit. A likely impact of the high indebtedness of Australian households is that the current economic downturn will be deeper and more prolonged than it might have been otherwise and/or more intensive policy intervention will be required to manage it. Household consumption is unlikely to recover fully until household balance sheets have adjusted to reflect expectations of future growth in household income and wealth. There has been much speculation on the probable impact of the recession on house prices. Although house prices fell, on average, by around 7 per cent in the year to the March quarter of 2009, the falls were concentrated in particular sectors of the housing market (that is, largely in wealthy suburbs and mostly in Sydney and Perth).5 The large-scale deleveraging of household balance sheets would place downward pressure on house prices, but the general undersupply of housing in Australia, exacerbated by population growth and immigration, exerts a braking effect. Certainly, an oversupply of housing, on the scale of that evident in the United States, is no part of the Australian economic landscape. Household indebtedness has increased sharply on a range of measures since financial deregulation. Higher indebtedness has raised the risks and returns for households, as well as for the macroeconomy. Households gained as they were freed from artificial credit rationing to better smooth their consumption over time and to leverage themselves into tax-preferred investments, like owner-occupied housing. However, the current recession has exposed the risks inherent in higher household leverage, including sharp reductions in household wealth as asset prices (including house prices) have fallen, and falling household incomes in the wake of rising unemployment. The effect of financial deregulation has been as predicted; namely, higher expected returns to households and the economy accompanied by higher risk. Consumption is more responsive to adverse economic conditions when household debt levels are higher but, by the same token, monetary policy is more effective at countering shocks to household income. Recent sharp falls in interest rates have supplemented household income substantially, mitigating the impact on consumption of falling income and falling asset prices.