3
Financial deregulation and household debt
In the post-war decades, the political elite in Australia, along with its confrères in other countries who had experienced the Great Depression, was acutely aware of the costs of Depression and determined to avoid a repeat. Depressions are the great destroyers of full employment and rising living standards, so the avoidance of Depression was congruent with these two other great aims of post-war economic policy — yet not always completely.
There were several occasions when Australia’s policy-makers did not hesitate to risk full employment when they considered this was necessary to avoid a Depression. In terms of the European Commission’s bellwether indicators of economic crisis, these occasions arose when Australia was in danger of over-borrowing from overseas. The medicine was simple: a lull in the growth of disposable incomes to reduce the demand for imports and hence reduce overseas borrowing. This was a thoroughly coherent response to a threat of over-borrowing that derived from a fall in the terms of trade. Effectively, the dampening of demand by fiscal and monetary policy speeded an adjustment that would otherwise have been at the mercy of slow and potentially disruptive changes in prices and real wage rates.
The 1970s were a perplexing time for those Keynesians who had failed to incorporate capacity limitations and the potential for cost inflation into their systematic thinking. Cost inflation broke the rule of thumb adopted by Treasuries in the post-war years, which was that any threatened incresase in unemployment should be addressed by expanding demand. This could be done without raising the threat of inflation, because the spare capacity inherent in unemployment would prevent price rises; conversely, the threat of inflation was to be addressed by dampening demand, which could be done without risking a risie in unemployment, because the over-utilisation of capacity meant that there were enough jobs to go around even if demand was reduced. Cost inflation spoilt this simple trade-off, and its control became the economic-policy issue of the decade.
Among countries on the American side of the Iron Curtain, a split developed between those that controlled cost inflation by corporatist means, centring on high-level negotiations which included trade unions and was backed up by qualitative financial controls, and countries that relied on the revived neo-liberal approach which centred on breaking the power of trade unions. In its arbitration system, Australia had an institution that could readily have been adapted to control cost-inflation by corporatist agreement, but two important groups were blind to this opportunity: the big-business elite, for whom the neo-liberal United States was the great exemplar; and the new generation of economists returning from initiation into the wonderful models of neo-liberal economics, primarily in American graduate schools. Australia lurched into neo-liberalism, not in one decisive event but in a series of decisions taken by governments both Labor and Liberal/National. The relevant governments called this process macroeconomic and microeconomic reform, and gloried in it.
This transition was only possible because the memory of the Great Depression had receded. The forgetfulness reached the point, in the United States, if not in Australia, where neo-liberal economists claimed that, because (by assumption) in General Equilibrium there could not be a Great Depression, the Depression must have been caused by government intervention. Suffice to say that when countries adopted neo-liberal policies, they generally dropped their guard against the threat of economic crisis. As discussed in Chapter 1, this nonchalance was associated with an increase in the frequency of national economic breakdowns such that, by 2010, enough crises had occurred to allow statisticians to identify bellwether indicators of the onrush of economic catastrophe. The authoritative work by the EC that we have already discussed identified ten such indicators, five of which are currently flashing red for Australia. These flashing indicators fall into two groups: the first group is primarily concerned with domestic debt, and the second with international debt. It would be irresponsible not to take these indicators seriously and to ask how they are related to the adoption of neo-liberal policies, particularly financial deregulation.
Private-sector debt
The three EC indicators that raise the alarm over Australian private debt are the rate of private-sector credit flow, the level of private-sector debt, and the rate of increase of house prices.
Despite its prominence in Australian media commentary, the government debt indicator is not flashing red. Neo-liberal economics provides a simple reason for this concentration of commentary: government borrowing is under government control, whereas private debt results from market activity, which by assumption guarantees equilibrium. It does not occur to neo-liberals that decisions made in more-or-less competitive private markets could add up to macroeconomic danger, just as it does not occur to them that borrowing by government is not always dangerous.
Fortunately, the EC has rid itself of these delusions, and selected its indicators of macroeconomic imbalance by empirical analysis of experience across the world. It has included rapid private-sector credit flow (the total of new borrowing by households and businesses as a percentage of GDP) among its indicators of impending crisis, because of its track record as a warning indicator. Rapid credit growth is associated with poor allocation of credit. For Australia, the indicator remained well below EC alarm levels during the post-war period; it may have risen above this level in the late 1980s,1 and it definitely moved above the EC alarm level in 1998, with a further rise in 1999. The next years of alarmingly high credit flow were the six years from 2003 to 2008. The Global Financial Crisis caused a pause, but dangerous levels resumed in 2014 and 2015.2
The second indicator of private-sector debt, which cumulates credit flow, is the level of household and business financial liabilities in relation to GDP. There is recent evidence that an excessively large financial sector is associated with a low economic growth rate, and very little doubt that high private-sector debt increases both vulnerability to financial crisis and the intensity of any crisis that occurs. In Australia, strictly comparable time series are not available for the post-war period, but such numbers as are available suggest that private-sector debt, though gradually increasing, remained well below the EC alarm level during the post-war period, but began to rise with deregulation. The indicator breached the danger level in 1998, and ever since has been above it. This increase was almost entirely due to an increase in household debt, which rose from 53 per cent of GDP in 1992 to 134 per cent in 2015, compared with an increase in the debt of incorporated businesses from 105 per cent of GDP to 110 per cent. Since 1992, though not necessarily in the first decade of deregulation, the increase in private-sector debt has been chiefly due to an increase in household-sector debt.
It has long been observed that booms in asset prices frequently turn out to be bubbles that, when they burst, generate financial catastrophe. The EC indicators zoom in on booms in house prices. The commission does not deny that booms in other asset prices can be harmful, but housing-price booms are particularly so. The wide spread of house ownership means that a bubble in house prices directly affects a much higher proportion of the population than does a bubble in share prices — since financial deregulation, Australian share prices have twice fallen by 40–50 per cent without precipitating a general economic crisis.3 By contrast, and in line with the alarming increase in household debt, house prices have risen with no comparable shakeouts, and are now the cause of considerable complaints focussing on poor affordability. Average house prices adjusted for the general level of inflation, having risen by around 1.4 per cent a year during the 1980s,4 increased by more than the EU alarm threshold of 6 per cent a year in several bursts: 2000, 2002–03, 2007, 2010, and then each year from 2013 to 2015.
In the post-war period, Australia accordingly remained below the EC alarm level signalled by its three private-debt indicators, but definitely rose above them in the late 1990s, around fifteen years after the main phase of financial deregulation. The rise in three indicators related to private debt was not, therefore, an immediate result of financial deregulation. However, a relationship appears once we explore further.
The first five years of deregulation
Touted as an epochal reform that applied neo-liberal principles to Australian policy, financial deregulation relaxed previous constraints on the ability of the banks to exploit market opportunities to increase their income-earning assets. Though their capital adequacy was still regulated, this was initially no constraint, thanks to the unlocking of funds previously sequestered in low-interest loans to the Commonwealth government. The banks also gained unregulated access to overseas lenders, and so long as the returns they received from domestic lending exceeded their costs of overseas borrowing, they could turn a profit on domestic loans financed by borrowing overseas.
Once freed from constraint, the banks rushed to add to their loan books, arguing that the more loans they could make, the greater would be their interest receipts and the greater their profits. Reinhart and Rogoff remark that a banking crisis tends to follow within five years or so of financial deregulation. The United States observed this rule; its ‘savings and loan’ banking crisis followed deregulatory moves in the 1980s. Australia was no exception. From 1984, Australia’s newly deregulated banks scurried to make loans to ‘entrepreneurial’ companies and to developers of commercial real estate — not, at that stage, to households, because they calculated that high administrative costs would restrict the profitability of small loans.
The 1987 crash of various global stock markets, including Australia’s, revealed that the large loans to entrepreneurial companies were imprudent, but it was not till the 1989 crash of the commercial property market that the banks found themselves in difficulties from bad debts. Right on time, five years after deregulation, ‘the 1990s began with the banking industry experiencing its worst losses in almost a century’.5 This banking crisis was associated with a dangerous level of overseas borrowing and a recession that was severe but stopped short of catastrophe. The banks wobbled, but survived.
Mortgages and equity withdrawal
In the early 1990s, the Commonwealth government was at its wit’s end as to how to promote recovery from the recession. After all, neo-liberal theory assumes that recessions do not happen. The authorities attempted to return to the post-war response of increasing demand by borrowing to finance public infrastructure investment, but found that the remaining public-sector infrastructure owners did not have sufficient shovel-ready projects available to make a quick difference. They therefore resorted to encouraging households to add to demand. The partial deregulation of gambling increased household expenditure at the expense of saving, but the measure that really raised consumption was increased bank lending to households.
Probably without realising what they were doing, Commonwealth governments in the 1990s caught themselves up in policy inconsistency. Blinded by neo-liberal economics, they promoted National Superannuation to increase household savings, without making any provision to ensure that these savings were invested in projects of national development. Instead, the governments came to depend on consumption to maintain demand. As quickly as households accumulated superannuation savings — indeed, quicker — they accumulated bank debt, with the net result of a reduction in the savings rate. This double accumulation suited the finance sector just fine.
The lesson that the banks took away from the 1990 recession was to avoid making loans to entrepreneurial businessmen. Rather than rein in their lending, they switched to an emphasis on household mortgages. Computers had reduced the administrative cost of small loans, and the banks were comfortable that these loans were secure, for three reasons. First, their borrowers were gaining assets that allowed them to save on rental costs and hence could readily be serviced, provided loans were restricted to borrowers with reasonable earnings prospects. Second, population growth generated strong demand for new houses, and the banks could argue that mortgage lending met an urgent social need. This argument justified the Commonwealth government in its encouragement of mortgage lending through its taxation system, particularly by allowing negative gearing, and also through its National Superannuation scheme, which promised lump-sum benefits to retirees that could be devoted to debt repayment. Third, the banks felt secure because, if all else failed, the properties provided collateral.
However, what seemed prudent and profitable to the banks and was convenient for the government turned out, with hindsight, to have been foolish policy from a broader macroeconomic point of view. Up until deregulation and indeed for some years thereafter, aggregate lending for housing had been limited so that the demand for new dwellings roughly equalled the capacity of the urban-development and construction industries to supply new conveniently located dwellings. When mortgage lending took off in the mid-1990s, it raised the demand for metropolitan housing to the point where it seriously outran supply, resulting in rising prices.
The limited production of new housing was no fault of the construction industry, which stood ready and willing to build. Instead it reflected a failure to create new house sites of acceptable quality. The quality most important in a house site is ready accessibility to employment. By 1995, the fringe of Australia’s major cities — the places where new house sites are most readily created — had moved beyond a reasonable travel-time distance to the centres of metropolitan employment, so that fringe houses were less and less adequate as substitutes for existing, better-located housing. Coincidentally, the decline of manufacturing industry (which was, at least partially, a result of neo-liberal policies) meant that jobs were no longer decentralising outwards from the metropolitan centres at the rate they had been in the post-war period. Instead, and in defiance of the American precedents on which the state governments in charge of metropolitan transport investment were relying, employment (and especially high-productivity employment in what became known as the knowledge economy) was re-concentrating in the city centres. It took decades for governments to realise that mass commuting from the metropolitan fringes to the city centres requires mass transit — the very public transport that neo-liberals wanted to scrap because it ran at a loss.
There are three main strategies to increase the supply of accessible housing. The first is to decentralise jobs to suburban centres and provincial cities; the second is to redevelop brownfield sites in already-accessible locations; and the third is to invest in fast mass transit connecting greenfield housing developments with employment locations. This is not the place to discuss their relative merits — in fact, there is room for all three strategies — but it is important to emphasise that each strategy requires considerable advance planning and investment; all three are slow-acting, hence the inability of the housing market to meet the avalanche of mortgage finance with an increase in supply.
The result was an increase in established house prices (or, more precisely, an increase in the price of inner- and middle-suburban land), which brought capital gains to established owners, frustration to aspiring owners, and a crisis of housing affordability. The increase in mortgage lending — in household debt — accordingly financed rising prices rather than construction. These rising prices in turn persuaded established home owners that it was safe to increase their consumption expenditures — either directly, by taking out a mortgage against the enhanced value of their home, or indirectly, when retirees sold out, harvested their capital gains, and shifted to lower-priced retirement regions.
Home-owning households were also encouraged to increase their mortgage debts by the lump-sum promises of national superannuation; within the limits of their expected lump sum, they could be persuaded that it was safe to finance consumption from a mortgage. These increases in consumption outweighed the decreases as home-buying households tightened their belts to pay high mortgage costs.
Mortgage-financed increases in consumption constitute the phenomenon of equity withdrawal. More formally, equity withdrawal comprises household debt incurred over and above the level required to support new investment in dwellings, unincorporated businesses, and not-for-profit enterprises. Looked at from a macroeconomic point of view, because it is not balanced against new investment, it is easy to see that equity withdrawal is used to finance consumption. It underlies growing dwelling prices and ultimately supports consumption expenditure via household wealth effects. By the end of 2015, household-equity withdrawal debt had reached 72 per cent of GDP, or over half of total household debt. The banks were not only lending to finance consumption directly (via credit cards and the like), but were doing so massively through their mortgage lending. Long and painful experience has shown that it is imprudent to lend (and to borrow) to finance consumption.
A major cost to the economy of the accumulation of equity-withdrawal debt lay in its counterpart on the liability side of bank balance sheets: increased vulnerability to overseas lenders. We consider this in detail below. However, even if attention is confined to the asset side of bank balance sheets, it was noticeable that the surge in lending to households crowded out lending to governments. At the end of 2015, total public-sector net debt stood at 28 per cent of GDP, while that part of total household debt which derived from equity withdrawal represented 72 per cent of GDP. If over the last decade just 20 percentage points of the equity-withdrawal debt had been transferred to the public sector and spent on infrastructure investment, the capital stock of the economy would have expanded by approximately $300 billion in 2015 prices.
Empirical estimates of the annual increase in GDP from an additional dollar of public-sector capital stock (technically, the marginal product of infrastructure capital) have generally been found to lie between 20 and 70 cents, with transport-infrastructure investment towards the higher end of the estimates. Even if a conservative marginal product estimate of 30 cents is assumed, national GDP would have been 6 per cent higher by 2015. The public-sector debt-to-GDP ratio would have been no greater than 45 per cent — well within the EC safe limit — and most likely significantly less, given the opportunity to use the extra tax revenue from the additional growth in GDP to pay down debt. The total private- and public-sector debt-to-GDP ratio would have been less than its present level.
The ease with which we can construct this alternative scenario, in which continued quantitative regulation of the finance sector would have resulted in more rapid growth of GDP than that actually achieved, points to a failure of financial deregulation, and of the neo-liberal reforms more generally, to deliver on one of their core promises. This was the promise of a rapid growth in productivity.
Productivity
It is almost a tautology that high-income countries have high productivity; that is, that they extract high incomes from their available resources. The concept of productivity concentrates on outputs (income) in relation to inputs (labour, capital). Capital is exceedingly heterogeneous and difficult to convert from a financial amount into a measurable quantity of physical inputs, whereas labour, while admittedly very diverse, can be measured in such terms as workers employed or hours worked. Productivity analysis has accordingly concentrated on the ratio of outputs (generally valued as income generated) to labour inputs.
The neo-liberals promised that their reforms would revive growth in productivity after the doldrums of the 1970s. Tax cuts would sharpen incentives to work and increase productivity, particularly at the top end of the income distribution, while welfare cuts would force bludgers to become productive workers. Tariff cuts would intensify overseas competition and force Australian trade-exposed businesses to lift their game. Privatisation would force government-owned businesses to disgorge hoarded labour, raising productivity. And financial deregulation would ensure that the flow of savings was invested so as to maximise productivity.
There is no doubt that financial deregulation brought one major increase in productivity, measured as income generated per person employed. This was in the finance sector itself. From 1984 to 2004, labour productivity (income in constant dollars generated per employed person) in the finance sector grew by 5.7 per cent a year, way in front of all other industries and well above the national average rate of 1.6 per cent. Even when the rate of growth of productivity declined from the mid-2000s on, the finance sector remained ahead.6 These increases were no chimera; the incomes so generated raised bank profitability and bank executives’ rewards. According to the national accounts, the share of the finance sector in locally owned private corporation gross profits (defined as consumption of fixed capital, plus undistributed income, plus dividends payable) rose to 28 per cent by 2015, well above historic benchmarks.
Underlying this increase in profitability, the size of the Australian financial sector grew from 5.8 per cent of nominal GDP at factor prices in 1992 to 9.3 per cent in 2015, with a similar change in real terms. Had the authorities succeeded in curbing equity withdrawal, this increase would not have occurred. The banks would not have become as fantastically profitable as they appear to have been, at least until their accounts are adjusted for the final impact of their imprudent domestic loans and overseas borrowings. In so far as the prosperity of the years since 1992 was underpinned by equity withdrawal, incomes would also have been less — but how much less would depend on how the economy was managed, since debt accumulation is not the only way to keep an economy moving.
By contrast with Australia, most Western European economies eschewed equity withdrawal as a driver of economic growth. In these countries, the size of the finance sector has remained stable at between 3 and 5 per cent of GDP. Norway, which, like Australia, had to manage a large-scale resource-development boom over this period, actually experienced a decline in the share of the financial sector in real GDP to 4 per cent or less. It not only managed to slim its finance sector; it also managed the over-valuation of its exchange rate to minimise the adverse effects of its resource boom on its non-resource industries, and maintained full employment to boot.
Analysis by the Organisation for Economic Co-operation and Development (OECD) showed that, on average, for each 1 per cent per annum growth in employment of the finance sector across 21 OECD economies between 1980 and 2009, the national productivity growth rate declined by at least 0.3 per cent per annum. Figure 1 indicates a similar relationship for Australia, where the negative relationship between national productivity growth and the rate of growth of finance hours worked is also -0.3.
Financial deregulation has dampened productivity growth in several ways. By reducing the care given to the management of small-business loans, it has missed opportunities and abolished a trusted source of financial advice to small business. By raising land prices, it has raised costs and reduced the competitiveness of land-using trade-exposed industries. By encouraging equity withdrawal, it has reduced the flow of savings available for the domestic finance of investment. By dampening government borrowing, it has curbed infrastructure investment.
In conjunction with the other neo-liberal reforms, financial deregulation has redistributed capital and labour resources away from high-productivity sectors to low-productivity sectors. It has redistributed labour away from the production of tradeable and high-technology goods and services towards low productive activity such as property, retail, and accommodation services. It has diverted capital resources from investment in product development, equipment, and industrial capacity, and towards commercial and residential buildings. By raising finance-sector profits and executive remuneration, it has contributed to the increase in inequality of income and wealth. These failures underlie the increased vulnerability of the Australian economy as measured by the EC indicators of private-sector debt.
Figure 1: Australia: Finance hours worked growth versus national productivity
Consumption as a source of demand
Insofar as the immediate purpose of debt accumulation is to add to purchasing power and thus to increase expenditure, one would expect an increase in debt to be associated with an increase in GDP. This relationship is measured by the ratio of the change in annual nominal GDP to the change in total private-sector debt. Between 1992 and 2012, this ratio was reasonably stable around 0.36, meaning that $3 of additional private-sector debt was associated with a $1 increase in nominal GDP. However, over the last three years the ratio has fallen to 0.18, so that $6 of additional private-sector debt is associated with a $1 increase in GDP.
Australia seems to be approaching the stage where debt accumulation is no longer effective in driving GDP. In other words, while in the 1990s and 2000s the boom in mortgage lending brought some macroeconomic benefit, this is now disappearing, leaving just the risk of crisis inherent in excessive debt.
As we have noted, the Commonwealth government condoned the surge of debt-financed consumption as a way of recovering from the 1990 recession. Given the political and timing difficulties of reviving demand through infrastructure investment other than roadbuilding, the tactic served its purpose. However, it now serves to accumulate debt without adding much to demand. Its time has passed.
The picture darkens further when we move to our last group of EC indicators: those dealing with overseas debt.