Menu Close

Ken Henry: why Australia’s non-mining sector will continue to struggle

There is no feasible set of adjustments around labour reform or productivity enhancements that will offset the loss of international competitiveness being experienced by the non-resources sector.

A new approach to tackling the structural effects of Australia’s record terms of trade on non-mining industries is needed from both business and government, argues one of Australia’s most eminent economists, Dr Ken Henry.

In a speech to the Business Symposium of the Australian Conference of Economists at Victoria University in Melbourne, Dr Henry has argued Australian businesses outside the booming resources sector will continue to face falling international competitiveness and options such as “offshoring” will need to be considered.

Dr Henry argues the debate around lifting Australia’s productivity and our international competitiveness has become confused, with a common but incorrect belief that growth in the resources sector can off-set weakness in the non-resources sector.

“The truth is, of course, that if the resources sector were to grow even faster, then the non-resources sectors must grow at an even slower rate,” Dr Henry says.

Dr Henry is Special Adviser to the Prime Minister responsible for leading the development of a White Paper on Australia in the Asian Century. He is part-time executive chairman of the Institute of Public Policy at the Crawford School of Public Policy at ANU.

Read the full transcript of his speech below.


Australia’s geographic location and abundant natural resources have been key determinants of its economic structure since human settlement. For as long as any of us can remember, Australia has been a significant exporter of agricultural products and bulk commodities, especially iron ore and coal, a feature that has distinguished us from most other countries at similar stages of economic development.

Trends in the international terms of trade have not always been favourable to Australia. Indeed, in the last few decades of the 20th century the popular view was that Australia’s particular natural endowments had consigned it to ever-falling export prices and ever-falling terms of trade.

That prospect helped motivate the policy reforms of the 1980s and 1990s. And those reforms paid substantial dividends in economic performance.

Even so, by the turn of the century, with the world in awe of impressive developments in information and communications technologies and the advent of electronic commerce, Australia’s natural resource endowments were considered by many to be a curse. Australia was being described, in disparaging terms, as an “old economy”, producing things that the rest of the world was increasingly finding ways of doing without. People were talking about the emergence of a so-called ‘weightless economy’ – an economy developing in the major European and North American centres of economic activity, drifting ever further away from Australia.

Today, in the Asian Century, Australia is often described as being “in the right place at the right time”, its geographic location and abundance of natural resources being seen as valuable assets.

Australia’s terms of trade are some 80% above their level at the turn of the century. In contrast, the real prices of information and communications technologies have continued to fall at a rapid rate. Both of these things can be explained, to a large extent, by the extraordinary growth of China as a manufacturing powerhouse, with a thirst for raw materials that Australia has in abundance.

Abundant as they are, Australia’s natural resources will not last forever. That observation warrants serious discussion, but it is not my focus today.

Even if it turns out, in the broad sweep of history, to be transitory, this terms of trade boom, which really kicked off in late 2003, has been with us for a while; long enough for it to have prompted a number of serious debates. Most obvious is the concern that the associated appreciation of the nominal exchange rate is producing a multi-speed economy in which mining and mining-related construction activities grow strongly while the rest of the economy stagnates or recedes.

Commentators today point to a decline in what they refer to as Australia’s “international competitiveness”.

This concern of a loss of “international competitiveness” is worth saying something about, and that’s what I’m going to concentrate on today.

“International competitiveness” sounds like one of those things that should always be encouraged, whether the terms of trade are rising or falling. And, indeed, in the 1980s when the policy concern was very much one of declining terms of trade there was intense interest in policy initiatives that might enhance Australia’s international competitiveness.

Yet, as economists, we also know that when the terms of trade are rising, especially if they are rising rapidly, a decline in international competitiveness is an expected part of the macroeconomic adjustment process. But that doesn’t mean that we consider international competitiveness to be something to be discouraged. And we are definitely in favour of anything that contributes to international competitiveness by lifting productivity.

No wonder the debate about these things has been a little confused. Hopefully, this conference will be doing something to reduce the level of confusion. We’ll see. In the meantime, I would offer the following perspective.

Let’s take, as our starting point, nominal unit labour costs. For a firm, or an industry, these can be expressed as the ratio of the nominal wage rate (expressed in $A) to the marginal product of labour. If labour is the only variable factor of production the nominal unit labour cost thus defined also represents the marginal cost of production for a given nominal wage rate.

If we multiply this ratio by the nominal exchange rate we have an expression for the marginal cost of production expressed in foreign currency units. This latter expression provides a summary measure of what most people seem to have in mind when they talk about the “international competitiveness” of a firm or industry.

Defined the way I have, the international competitiveness of a firm or industry falls when its marginal cost of production, expressed in foreign currency units, increases. This can happen because of an appreciation of the nominal exchange rate, an increase in the nominal wage rate, or a decline in labour productivity.

In recent years, the nominal value of the $A has appreciated strongly, aggregate wages have grown at an above average rate and aggregate labour productivity growth has slowed to a rate significantly below its historical average. Driven by the consequent loss of international competitiveness, the sectoral pattern of output growth has exhibited a marked divergence as between the mining (and mining-related construction) sector and other sectors, and as between the resources-rich states and other states. None of these things is at all surprising; and, indeed, all of them have been predicted for at least the past six years.

Even so, as they have emerged, they have generated considerable angst, and that angst is increasing.

Australia is frequently labelled these days a “low productivity, high cost” country. Business groups and others are calling on governments to do something to lower costs and boost productivity; and by those means, restore Australia’s international competitiveness.

Now I don’t want to discourage anybody from calling on Australia’s numerous governments to engage in microeconomic reform, so I’m sure what I’m about to say won’t be misinterpreted. But I also hope it will be understood. I just want to be sure that we all understand what might be the consequences of some of the things we are calling on our governments to do; and, in particular, whether those things would be enough to achieve what we say we want.

I’m going to illustrate some issues with a very simple numerical model. This model will be considered simplistic. So why do I bother? First, most of the discussion of these issues in Australia today is qualitative. In that sort of discussion it’s hard to know what importance should be attached to various possibilities. For example, what has been the relative importance of poor management and poor labour laws in explaining Australia’s poor productivity performance? Numbers would be useful. Second, though, I don’t want to draw some numbers out of an impenetrable black box. The Australian community of economists has a lot of those black boxes. I’ve constructed a few, myself, over the years. Rarely does their use shed light on issues, for the simple reason that they are generally poorly understood. In any case, despite their complexity, they are no more realistic than the simple model I’m going to use here today.

One other qualification: The numbers I am going to report are not forecasts, predictions or projections. I am using them simply to put rough orders of magnitude on some things.

In this simple model, we have a trade-exposed resources sector and another trade-exposed sector that competes with the resources sector for workers. We will label that second sector “non-resources”.

Both of our sectors utilise Cobb-Douglas production functions – the simplest form of production function used in empirical work, and which is familiar to all students of economics. The Cobb-Douglas form has some useful properties, including diminishing returns. Both sectors adjust their scale of production to ensure that their marginal cost of production matches their domestic price. In both sectors, domestic prices are simply “world prices” divided by the nominal exchange rate.

The resources sector is more capital-intensive; let’s suppose it has a capital share of 70%, and that the non-resources sector has a capital share of 30%. We will suppose that any gains in multifactor productivity, in either sector, emerge as Hicks-neutral technical progress in that sector; that is, each of the production functions has a scalar sitting out the front of it that can be manipulated.

Starting from an initial equilibrium in which the non-resources sector employs 95% of the labour force available to the two sectors, suppose the world price of resources products increases four-fold. This is not an extreme illustration. In the past decade, the $US price of Australia’s iron ore exports has increased more than ten-fold, while the $US price of Australia’s coal exports has increased more than five-fold. And, by the way, these two industries today contribute more than a third of Australia’s total exports.

We will consider some of the implications of this price shock. In what I will call the “base case”, there is no change in multifactor productivity, or capital utilisation, in either industry; there is no change in the world price of “non-resources”; and there is no change in the total stock of labour available to be shared between the two sectors.

So what happens in this simple model when world resources prices increase four-fold?

The first thing to note is that, with no change in the wage rate, or nominal exchange rate, the resources sector’s demand for labour will increase by about 625 per cent. Yes, even though this sector is highly capital-intensive, that is the order of magnitude.

That sort of increase in labour demand can be expected to have some impact on wages. The size of the impact depends upon the mobility of labour between the two sectors. If workers are perfectly mobile, then it turns out that a nominal wage increase of a little more than 10% will clear the labour market of excess demand.

Of course, the assumption of perfect labour mobility is unrealistic. Suppose, instead, that the shock to the world price of resources opens up a 50% wage premium in that sector. Then the labour market will clear if wages in the resources sector increase by about 55% and wages paid in the non-resources sector increase by about 5%. Obviously, in order that no reallocation of labour occurs, wages in the resources sector would have to quadruple – leaving wage rates unchanged in the non-resources sector.

These wage increases should be thought of as multi-year adjustments, occurring on top of trend wage developments. How many years? I don’t know.

In what follows, just to keep the explanation of things as simple as possible, I’m going to suppose that labour is perfectly mobile between sectors, implying the same nominal wage increases in resources and non-resources.

Now, the central bank might not be comfortable with the implications for inflationary expectations of the sorts of wage increases I have identified, coming, as they do, on top of trend wages growth. For that reason, and others, the nominal exchange rate will appreciate. Let’s suppose it appreciates by a little over 10%. Then, in order that the labour market clears, the wage rate doesn’t have to move at all. That is, the currency appreciation is just sufficient to choke off the excess labour demand associated with a quadrupling of the world price of resources.

None of you would have been surprised to learn that in this simple model, the nominal wage rate and nominal exchange rate are perfectly substitutable variables. That is, changes to the nominal exchange rate have absolutely no impact on any real variable, like employment levels, labour productivity and output; they simply affect the magnitude of the nominal wage increase required to clear the labour market.

[Of course, this is not true if labour is not perfectly mobile between sectors. In that case, employers in the non-resources sector would not have to match wage increases in the resources sector in order to hold onto their labour, but they would still be made uncompetitive by an increase in the nominal exchange rate, shedding labour that would become unemployed.

To summarise thus far: In the base case, in which the world price of resources quadruples, and there is no change in multifactor productivity, or capital utilisation, in either industry; and no change in the world price of non-resources; and no change in the total stock of labour available to be shared perfectly elastically between the two sectors, either the nominal exchange rate or the nominal wage rate has to increase by a little more than 10% in order that the labour market clears.]

Note that in the real world, the nominal exchange rate appreciation consequent upon the resources boom has been a lot bigger than 10% – several multiples of it, in fact.

What about the impact on real variables like employment, productivity and output in our base case? Well, they’re big.

Employment in the resources sector increases by more than 500% and employment in the non-resources sector falls by almost 30%. So the resources sector now employs about 30% of the total labour available to be shared between these two sectors, with non-resources’ share falling from 95% to about two-thirds.

Output expands by almost 75% in the resources sector and contracts by about 20% in the non-resources sector.

These are big numbers, but they make sense. If the resources sector receives a quadrupling of its price and can access labour from another sector of the economy, that other sector is going to be hit heavily, even if it initially employs nearly twenty times as much labour as the highly capital-intensive resources sector.

Because the stock of available factors of production cannot expand elastically, an expansion of activity in the resources sector necessarily crowds out activity in the sectors with which it competes, internally, for those inputs. This crowding out occurs through increases in the exchange rate and wage rates. Either of these would appear as a loss of international competitiveness in the trade-exposed non-resources sectors.

At least in respect of the direction of change, if not magnitude, all of this will make sense to you. But I would ask you to reflect on the number of occasions, in the past several years, you have heard commentators saying that the key macroeconomic question for Australia is whether the resources sector can grow fast enough to offset weakness in the non-resources sectors of the economy.

The truth is, of course, that if the resources sector were to grow even faster, then the non-resources sectors must grow at an even slower rate.

Our simple numerical model permits us to explore some other things of interest. In the base case, labour productivity in the resources sector falls by more than 70 per cent; but it increases, by a bit more than 10 per cent, in the other sector. Economy-wide, average labour productivity falls by about 10 per cent.

Looking at the resources sector first, it has had a 300% increase in price. So it can afford to expand production to the point where marginal costs are 300% higher. Now recall that the marginal cost of production is simply the ratio of the nominal wage rate to the marginal product of labour. This ratio must increase by 300%. If the numerator increases by only 10%, most of the increase in marginal cost is going to come from falling labour productivity. The new level of labour productivity will be a fraction of its initial level. This fraction is about 1.1 divided by 4, which is less than 0.3, implying a reduction in labour productivity of more than 70%.

In the non-resources sector, since the output price has not changed, neither can the marginal cost of production. So if the nominal wage rate has increased by about 10%, labour productivity must also have increased by about 10% in that sector.

In both sectors, firms are moving along their respective marginal product of labour curves. Recall that, for the moment, we have not allowed the capital stock to change in either sector. The resources sector expands production by hiring more workers. As it does so, it moves down along its marginal product of labour curve. The non-resources sector has to cut back production and shed labour. As it does so, it is moving back up along its marginal product of labour curve. It is this move, back up along its marginal product of labour schedule, which restores the non-resource sector’s level of international competitiveness. It becomes more competitive by shedding labour.

Obviously, a lot of people have other ideas about how to restore that sector’s international competitiveness. I’ll come to some of those shortly.

First, note that capital invested in the resources sector has become a lot more profitable. Indeed, its marginal product has increased by more than 50%. Let’s say it was earning normal profits of $10 billion before the increase in the world price of resources. It would now be deriving, in addition to those normal profits, $16.7 billion of super-normal profits.

So we can expect an investment boom in the resources sector. Investment in the resources sector boosts labour productivity in that sector. But there is no feasible increase in the resources sector capital stock that would prevent its marginal product of labour from falling at all.

Of course, any increase in mining capital stock does mean an even larger reallocation of labour and an even higher nominal wage rate and/or nominal exchange rate.

Are there alternatives to a very large reallocation of labour between the resources and non-resources sectors of the economy?

One suggestion is that the aggregate labour force be expanded so that the increase in resources sector employment doesn’t have to come at the expense of non-resources. What sort of increase would be required? The answer is that the labour force shared by these two sectors would have to expand by about 140%, with all of the additional labour going into the resources sector. I have to say that one does have to wonder how that sort of workforce augmentation could be achieved, especially since the mining boom, and its extraordinary demand for labour, is a world-wide phenomenon – with Australian miners competing for labour not just domestically, but with global competitors.

Another suggestion is to expand multifactor productivity in the non-resources sector. The required increase is larger, the larger the level of capital investment in the resources sector. Even if the resources sector capital stock doesn’t expand at all, a 300% improvement in the level of multifactor productivity in the non-resources sector would prevent it losing labour to the resources sector. The nominal wage rate – alternatively the nominal exchange rate – would also appreciate by 300%. That’s a big number. But it’s pretty straightforward arithmetic actually.

If the goal were to prevent non-resources output, rather than employment, from contracting, the position would appear more feasible: a multifactor productivity improvement of about 20%. Of course, even that is massive, given that it would come on top of “trend” changes.

So where does this leave us?

All of you will be thinking that the real world’s a lot more complicated than the simple model I have used here. Some of you might be thinking that that greater complexity means that the orders of magnitude of the adjustments don’t need to be as large as the numbers I have presented here. Maybe. I would simply point out that some of the adjustments that have occurred in the real world – including the appreciation of the $A – have been a lot larger than what this simple model would deem necessary.

I haven’t set out these numbers in order to dissuade anyone from arguing the case either for an expansion of the Australian labour force – through immigration or participation-enhancing policy reforms – or for a renewed focus on productivity enhancing policy reforms. Indeed, I consider there to be a strong case for all of these things.

But there is simply no feasible set of such adjustments that would reverse all, or even a large proportion, of the loss of international competitiveness that is presently being experienced by Australia’s trade exposed non-resources industries. The shock to resources prices has simply been so large that a considerable structural adjustment, including a reallocation of labour, is required. It will happen.

Some of you might be wondering if we might be “saved” from all of this structural change by a large global resources output surge that will drive down the world prices of iron ore, coal and so on. To you, I would say this: much of what will, over time, drive world prices lower is precisely the structural change currently underway in Australia. Lower world prices will be, in part, the consequence of structural change in the Australian economy, not the thing that prevents it. Of course, the output response of other resources producing economies will also be important here – but that has been the case through all of the past eight years as resources prices have moved ever higher. It is true that Australia is not the only economy experiencing substantial structural adjustment because of the re-emergence of the major Asian economies. But that is not the same thing as saying that Australia will avoid substantial structural adjustment.

Structural adjustment in Australia could be assisted by implementation of some of the productivity and participation enhancing initiatives being proposed in present debate. I would endorse, especially, those initiatives that target a more efficient usage of intermediate goods and services – in transport, logistics and, of course, in sourcing – including exploring opportunities for what, these days, is usually referred to as “offshoring”.

But much, much more is required. And, as in the case of “offshoring”, most of it simply cannot be delivered by government.

It is no exaggeration to say that, if Australia is going to navigate successfully the structural adjustment to the terms of trade shock presented by the extraordinary growth of China and others in the region, a new mindset will be required: a new mindset in government, certainly; but a new mindset in business and the broader community also.

Want to write?

Write an article and join a growing community of more than 181,000 academics and researchers from 4,921 institutions.

Register now