RBA Annual Conference – 1995 Discussion

1. Henry Ergas

Case studies of industry restructuring all too often fall into two literary genres. The first is the ‘Boy's Own’ school of analysis, full of villainous foreigners, visionary heroes, courageous gestures and ultimate triumphs – readers of Peter Roberts' column, in which the brave Aussie battler invariably wins the day, will immediately recognise the style I have in mind. At the other extreme, and especially well-represented in the writing of economists, lies the cautionary tale. Like fallen maidens, firms, industries and countries are shown to lose their way; and regardless of the efforts they might subsequently make, virtue once lost is almost impossible to regain – although, extreme penitence and liberal self-flagellation may help, and are in any case widely seen as being their own reward.

The papers presented in this session mercifully avoid these approaches and provide a useful basis for examining aspects of the Australian experience. The industries they cover differ greatly, as do the changes they discuss. Yet there are some common elements, and it may be helpful to bring them out. Viewed as a whole, they suggested a profound change in the way Australian business operate – one yielding greater efficiencies, but not at zero cost.


The BHP story highlights the old way of doing things. Well into the 1980s, insular and indifferent management, sheltered both from real competition in product markets and from the threat of take-over, made little effort to modernise plant and equipment, to cut costs, and gain new markets. Thus, in 1954, just as world steel trade was set to expand, BHP's Chairman, while noting that ‘unquestionably the current demand … exceeds our present capacity’, argued against expanding to meet export needs because ‘the cost of freight from Australia to export markets is such that our steel can only be exported at satisfactory prices when a steel shortage exists throughout the world. … Sound expansion can only take place at a rate in step with the overall home market’ (Syme 1954, p. 173). Thirty years later, similar views were echoed in BHP's evidence to the Industries Assistance Commission; and when the Commission queried the slow pace at which new technologies were being introduced into the company's steel making, it was told that ‘although [the company] acknowledged that its plant is old and does not embody the latest available technology … it (noted that) the adoption of modern technologies would not necessarily result in lower costs … In view of the discrepancy between the depreciation on written down plant relative to the cost of new plant, the adoption of the latest technology will not result in a reduction in costs in the short term’ – the clear inference being that it was consequently not worth doing (Industries Assistance Commission 1983, p. 114).

Management's ‘quiet life’ was strengthened by the compact with the labour force. The Award system took much of the job of management away from those whose job it was. Conflicts were accommodated, rather than resolved; conditions, once granted, became entitlements; and the frequent disputes were primarily a way of triggering a conflict-management system ‘essentially as conservative as the unions and the company, and all … appeared to be satisfied with the status quo; alternative methods of settling working conditions and disputes seemed unnecessary, and perhaps more to the point, unrealistic’ (Hughes 1964, p. 172).

The changes which occurred in the 1980s – the reduction in border protection, the liberalisation of capital markets, the moves towards increased flexibility in the industrial relations system – had two effects: they made the old way of doing things unviable; and they pushed firms towards a new way, in which an increased orientation to world markets (encouraged by the more realistic level of the exchange rate) has been accompanied by the search for greater flexibility and an enhanced ability to cope with the disequilibria which competitive markets always create.

This last point is worth emphasising. The changes we observe in firm behaviour do not only bear on the level of costs – they also involve the structure of those costs, and the effects of that structure on the firm's risk. In particular, increased competition – both in the product market and in the market for corporate control – has led firms in a wide range of industries to seek lower break-even points: to find ways of being less exposed to high fixed costs in an environment where demand and prices are less predictable. One result of this is great caution in capacity expansion: an enhanced incrementalism, in which alternatives are sought to large, lumpy and irreversible expansions in plan. BHP's recently announced investment in electric arc technology is a case in point, because it so clearly trades off higher marginal costs for substantially lower fixed costs.


The NAB case study also highlights the gains being made in productive efficiency – and leads especially directly to a consideration of the impacts on employees. Three points need to be made by way of background.

First, compared to its domestic rivals, the NAB has invested more and better in information systems which allow it to support and monitor outcomes both for individual employees and at the branch level.

Second, given these systems, the NAB has gone a fair way in restructuring operations so as to separate activities which differ in terms of the continuity of their work-flow, the nature of the physical assets needed to support them, and the skills they involve. Most recently, the NAB has, in South Australia and Victoria, separated all the back-office operations out of its branches and centralised them, while extensively redesigning the tasks involved. At the same time, all telephone enquiries nation-wide are now handled out of a single location at King Street in Melbourne.

Third, drawing on improved knowledge of work-flows and required skills, the NAB has substantially increased the flexibility of its resource use. Under its first Enterprise Agreement, part-time employment increased to around 20 per cent of overall hours; together with temporary employment, it is set for further substantial increases under the second Enterprise Agreement which has now been negotiated.

Viewed as a whole, these changes should make it far easier for the NAB to respond to market change; but there is one aspect of these changes which is especially interesting, in my view, and which I would draw to your attention.

This is the use of information systems to track the performance of individual employees against quantitative targets. In the past, many activities had no specific performance targets set for them – for example, the time taken by counter staff to respond and clear down a telephone inquiry. Over the past decade, task redesign (as the consultants call it) has involved specifying tasks and task sequences in great detail, translating these into formal, task-support systems, defining goals for each step these systems identify, and establishing robust (usually automated) processes for measuring performance relative to the goal set. Three consequences of this, which are still far from playing themselves out in Australia, are clear when one looks at banks or indeed, other service activities, in the US.

First, among clerical and sales staff, firm-specific skills have become much less important than general skills (such as computer literacy) – so that firms have far fewer incentives to hoard white-collar labour during downturns. This, in large part, accounts for the increased cyclicality of white-collar employment.

Second, given relatively objective measures of individual performance, poorly-performing staff are likely to be identified and if necessary dismissed far sooner than was previously the case, while the remuneration of those who remain becomes far more dependent on performance-related and incentive pay. The result is increased intra-skill-group dispersion in earnings.

Third, given that shirking can be directly detected and corrected, the middle-management role is very much more limited, and indeed, in its purely supervisory form, tends to disappear.

These processes are at work in virtually all white-collar activities; and although their impacts on productivity are difficult to measure for all the reasons given in the Conference papers, they clearly have major effects on the structure of earnings and more broadly on social relations. They can do a great deal to help firms improve performance and cope with more rapidly changing market conditions; but it may well be that some part of the firm's increased ability to cope with risk is being obtained by shifting that risk back onto employees.

Pacific Power

This brings me to my third and last case study, namely Pacific Power. Here the impetus for improvement has come not from the product market – at least, not yet – but from the changing demands of the shareholder. The paper does a very good job of assessing comparative productivity levels and I have little to add to the analysis it presents. I wonder if there is not more to be said about the nature of the challenge ahead – or at least about the challenge which Pacific Power will face, when it is fully exposed to competition.

Again, the US experience is illuminating in this respect. What has become increasingly important in determining performance in the electrical industry over the past decade is the capacity to manage risk. As commercial pressures have forced down reserve margins, difficult issues have arisen about how and by who the risks involved in capacity expansion will be borne. With utilities ever more unwilling to bear uncompensated risk, the failure to adequately resolve risk allocation problems is now seriously distorting investment decisions in the industry, greatly increasing the likelihood of severe power outages. Similar issues will need to be addressed in the Australian market, but will be very much complicated if a uniform approach is not found among the relevant States, notably to the treatment of stranded assets. Given the political commitment to liberalising this market, it is relative to the demands of this inherently more complex and risky competitive environment, rather than against the benchmark of other monopoly carriers, that Pacific Power's performance needs to be gauged.


Although the circumstances dealt with in the papers differ greatly, many similarities can be identified. The one I have stressed is the importance being given in industry not merely to reducing cost but also – and perhaps especially – to better managing the uncertainties inherent in competitive markets. The old social compact which characterised Australian industry – protection all around, undemanding management and owners, and an industrial relations system which accommodated conflict (rather than trying to resolve it) – no longer exists. But even the largest firms are having difficulty in creating a new one. For the thousands of smaller firms, the uncertainties and challenges are even greater – as are the costs of failing to adjust.


Hughes, H. (1964), The Australian Iron and Steel Industry 1848–1962, Melbourne University Press, Parkville.

Industries Assistance Commission (1983), Certain Iron and Steel Products, Report No. 232.

Syme, C.Y. (1954), Report of the One Hundred and Second Ordinary General Meeting, Broken Hill Pty Ltd, Melbourne.

2. Barry Hughes

This cluster of papers is best seen as an extension of Philip Lowe's painstaking attempt to flesh out, at the micro level, the macro productivity results. As I understand the state of play, macro research tells us that for the non-farm market sector:

  • labour-productivity (LP) growth was slow in the 1980s, but it is possible, though not certain, that the pace is better in the 1990s;
  • capital-productivity (KP) growth jumped in the 1990s; and
  • as a result of these two statements, it is easier to demonstrate a structural KP than a LP shift in the 1990s.

No doubt because of its wage connection, and its greater use in general debate, attention has been concentrated disproportionately on labour productivity. Throughout the period since the mid 1980s there have been expectations that a surge in labour productivity was just around the next corner, though in the event the path seems to have had more twists and turns than the Adelaide Grand Prix strip.

We can put down the relative labour productivity disappointment partly to impatience. Economy monitoring is often as exciting as watching the grass grow. Things take time to evolve, and some things take a very long time. But part of the disappointingly long time for the fruit to appear might have been due to earlier inappropriate expectations. What should we have expected about productivity growth? These observations go some way towards explaining events.

One fountain of change was the opening of the economy, and especially the rounds of tariff cutting. But this is a relatively recent development, with significant progress on tariffs really starting from the very late 1980s and early 1990s rather than the mid 1980s. Change might have been threatened in the earlier period, but that threat was not a new event. What was new this time was that the threat had enough strength to overcome the usual political opposition, a position that did not become self evident until the recession of the early 1990s. In any event the significant steps along the tariff cutting road were the statements of 1988 and 1990, with actual changes spread out way down the subsequent track.

Tariff cutting might have been expected to be as much a capital as a labour productivity story. What was strange about the old ‘fortress Australia’ policy was the proliferation of small-scale capital plants, with Premiers competing amongst themselves to secure ‘industries’ (never firms) for their own patch. Auto assembly was a classic 1960s/1970s result, with scattered single-shift plants in every mainland State, destroying capital productivity in the process. The most obvious waste to this immigrant's eyes was one of capital rather than labour, and it has been no surprise that when the global competition blowtorch was fired up the first input to be economised was expensive capital (through second shifts and other extensions of plant usage over the working year).

A second, related, wellspring of change came from industrial relations dynamics. The two previous points can be repeated. Despite the hype, the change in industrial relations did not get underway until the very late 1980s at the earliest. Accords Mark III, IV and V of 1986 to 1989 were about disguising real wage reductions by shifting the focus of debate onto future gains from productivity growth, and then about distributing centralised wage gains in return for promises to think about productivity changes (and clearing up award structures), rather than implementing more efficient practice at the workplace. Enterprise bargaining is a trend of the 1990s rather than the 1980s.

Again it is just as easy to see this bargaining as a capital than a labour productivity story – in extended hours of operation of coal mines, retail stores (the unmentioned flip-side of Lowe's account), banks, the hospitality sector and manufacturing. This is not to deny that some of the changes were planned to lift labour productivity; the point is that the labour changes were only part of the story (and possibly initially the minor part).

Labour productivity growth comes from both TFP growth (the object of bargaining change) and increased capital intensity. The period of the early 1990s when enterprise bargaining began to take hold just happened to coincide with severe recession, a slump in investment and stagnation of capital – labour ratios. In other words, through this period of the first half of the 1990s, whatever was being achieved by way of TFP (and there is some evidence of a structural lift) was being obscured in its impact on LP growth by an unusual absence of the contribution from capital intensity. Given this conjuncture, it is perhaps not surprising that it is difficult to find persuasive evidence of a first-half-1990s structural lift in LP growth. Only now, with stronger equipment spending, should we expect both LP proximate sources to be working together – an optimistic thought for the near future (yet another corner to turn with optimistic expectation).

The Papers

All of the foregoing should have cautioned us to expect long delays before obvious productivity gains showed up on the scoreboard. How do the papers fit and extend this story? They extend the story by pointing to special non-macroeconomic factors that depressed productivity in the 1980s.

Lowe's important stress on deregulation's impact on retail labour productivity (although he accepts too readily for my taste the proposition that measured sales value added was unaffected by extended hours – note that the saving ratio has fallen – and omits the capital productivity flip-side of the story). A similar story emerges from a study of hospitality sector labour productivity (again with potential capital productivity implications neglected).

Although the industry is defined out of the ABS ‘market’ sector, Alan Oster and Lawrence Antioch's discussion of financial sector trends could have told a parallel story. Retail banking deregulation led to new outlets being established in the 1980s, with traditional banks hanging on to their premises for fear of giving the new competition an easy entry. I do not know what should be allowed for the extra convenience of even more banking outlets (not much shoe leather), but this phase must have blunted, and perhaps even reversed, the labour-productivity benefits to be gained from the higher capital intensity of computerisation and ATMs. But it provides greater hope for the 1990s now that this deregulation detour is ending.

The New South Wales Treasury study draws attention again to yet another adverse special factor of the 1980s in the form of the consequences of hugely-disappointed energy demand forecasts strangling electricity TFP in that State as the large discrete additions of Eraring and Bayswater came on stream. There is hope for the 1990s both in the record to date, and in the extra flexibility in dealing with future capacity additions to be afforded by the interstate grid.

Of course, there ought to have been particular successes that would have helped offset these negatives. Steel is often cited as an example, but even starting from a poor position in 1981, Peter Demura's paper reminds us that physical labour productivity (tonnes per steelworks employee) grew only 26.9 per cent in the seven years to 1988. It was subsequently, especially in the period since 1991, that productivity surged. Things often take a long time coming to fruition.

Partly in this context, Peter Demura's paper was valuable also for its short account of the trials and tribulations of BHP's attempt to lift productivity in the Port Kembla slab caster. Following the failure of much capital expenditure and labour effort, output suddenly lifted 20 per cent in the two years after 1992. Demura attributes the dramatic breakthrough to learning by doing (properly through teamwork and what used to be called industrial democracy) rather than simply through capital expenditure. There seems to be a fertile field for in-depth case studies of this sort, untangling both the contributing factors to success and failure in productivity improvement and the time-scales that typically might be involved.

Adding in contributions from possible measurement errors (after Gordon and Lowe), some important new light on the dismal recorded productivity of the 1980s might have been generated by this Conference. We are reminded repeatedly of the old saying that ‘time is a device to stop everything happening at once’. In the productivity field, everything rarely seems to happen all at once, with events conspiring to slow gestation periods.

The Apple, Belchamber and Bowtell paper also talked at length about the people involvement in productivity promotion. Unions are now ‘for productivity gains and competitiveness’, but it is hard to discern what their role in all this is other than cheerleading from the sidelines. I went looking for an exposition of the positive contribution that teams might make (with unions having a potentially important role to play in team formation), but I came away disappointed from their paper. There is a long list of ‘motherhood’ statements, but no assessment whatsoever of whether the union involvement facilitated or hindered the process of change. There remains almost a virgin field for research in the area, following up the sort of mini-case study presented by Demura but, for the moment, the union-productivity connection remains largely a matter of politicians and journalists at 20 paces.

3. General Discussion

The discussion centred on three main issues addressed in each of the case studies:

  • the motivation for productivity reform;
  • the process of reform; and
  • the challenges to sustaining this process.

A common feature of the case studies is that reform was motivated by some crisis or by demonstrably poor productivity performance that jeopardised competitiveness. Enterprises were compelled to address the question: how did we get into this mess? It would appear that existing practices proved unable to deliver adequate productivity performance in the presence of shocks or the threat of increased competition.

It was noted that while initial responses related to cost cutting and rationalising resources, there were subsequent changes in the organisation of work and organisational culture that led to greater ‘production consciousness’. It was emphasised that this consciousness was not confined to enterprises. It was also recognised by the union movement which has undertaken reforms that complement those of enterprises. This represents a new form of collaboration, at the enterprise level, to achieve common goals of economic performance.

Despite these important changes, though, it was argued that much of the improvement in productivity observed to date might be described as recovery from ‘bad luck’ or ‘bad mistakes’. If so, the productivity benefits of microeconomic reform might, in fact, be small. A widespread view was that the next major change in productivity performance can only be achieved through technological advance.

Consequently, in the interim, it was questioned whether Australian enterprise will be able to achieve world best practice. Indeed, it was argued that Australia is sufficiently behind world best practice that we are a candidate for technological catch-up, in contrast to the messages of the Dowrick paper. In reply, it was proposed that microeconomic reform is in many ways a response to the threat of other nations catching up in levels of per capita income, so that the process of catch-up and convergence is a continual chase.

Finally, it was suggested that concerns about the performance of individual enterprises or industries relative to world leaders should be tempered with the knowledge that no country has a monopoly on world best practice: all countries have scope for improving productivity in some sectors. Best practice is, itself, perhaps the most mismeasured concept.