National Economic Review
National Institute of Economic and Industry Research
No. 68 October 2013
The National Economic Review is published four times each year under the auspices of the Institute’s Academic Board. The Review contains articles on economic and social issues relevant to Australia. While the Institute endeavours to provide reliable forecasts and believes material published in the Review is accurate it will not be liable for any claim by any party acting on such information.
Editor: Kylie Moreland
©National Institute of Economic and Industry Research
This journal is subject to copyright. Apart from such purposes as study, research, criticism or review as provided by the Copyright Act no part may be reproduced without the consent in writing of the relevant Institute.
Net benefits of mining expansion
Dr Peter Brain, Executive Director, NIEIR
The present study examines the net benefits to the Australian economy of a mining boom. In light of the changed circumstances that are likely to prevail over the coming years, extrapolation of past responses to mining expansion into the future suggests that there may be little headline net per capita additional benefit. The resource claims to meet the infrastructure and service demands of the increased population induced by the current episode of mining expansion will be presented in full, creating very difficult political and economic constraints, adding to those from climate change. The situation will be compounded by the national productivity growth tending to remain below historical trend levels. However, the resource expansion could be managed differently, to maximise its net additionality. Such management would include increased harvesting of resource rents and measures to increase the domestic content of mining investment.
In a previous article we argued that, because the mining industry produces standardised commodity products, the drivers of expansion in the industry are very different from those in industries which develop and market differentiated and branded products (Brain, 2012). The commodity-production nature of the industry means that bursts of expansion generally occur as a response to unexpectedly high mineral prices, although they can also take place as a response to the discovery of new low-cost resources.
The current Australian episode of high investment in additional capacity to produce iron ore and energy minerals was generated by a burst of high prices. Past experience is that such episodes induce sufficient capacity expansion to increase world supply and so bring prices back towards costs of production, ending the investment boom. An episode of mining expansion thus has several phases: an initial phase of normal production; a phase of investment and high construction activity induced by high prices; a phase of increasing output as additional capacity comes on stream; and, finally, the stabilisation of production, generally at a higher level of output but not necessarily at higher prices than during the first phase.
During the construction phase the mining industry makes major demands on the construction industry in the country where the investment takes place. In particular, the demand for labour is high during the construction phase but falls during the enhanced output phase, a characteristic which is generally true of investment–output sequences. This demand for skilled construction workers can be met in four ways:
- the existing workforce could be used more intensively, so that other construction activity is unaffected;
- labour could be diverted from other construction to mining investment;
- guest workers could be used from overseas, with workers obliged to return home after completion of the investment campaign; and
- immigration of workers with the appropriate skills could be increased.
Each method has its disadvantages.
- If mining investment is limited to that which can be undertaken by available labour, even working intensively, the amount of investment will be limited, perhaps to less than that desired by investors.
- If labour is diverted, other construction programs will have to be curtailed. The question is then whether the mining investment has higher priority than the curtailed investment.
- Guest workers have a habit of becoming permanent migrants, in which case the disadvantages of the immigration solution become relevant.
- Increased immigration raises the question of how to find work for the immigrants (or for the established residents they have displaced) once the construction boom is over and the demand for labour has subsided.
The demand for construction materials also rises during periods of enhanced mining investment, but this demand can readily be met from imports if for any reason there is a shortage of local supply. Indeed, the global strategies of mining investors often favour overseas sourcing, because this adds to demand in their home countries. It is mainly in regard to labour that an important analytical question arises. From a country’s point of view, who are to be regarded as the potential beneficiaries of a period of mining expansion: the population of the country as at the beginning of the expansion plus its descendants or the population as augmented by any induced migration?
Not only is it important to be clear as to the population relevant to assessment of the benefits of mining investment, it is important to be clear as to the metric of assessment.
Indicators of national economic welfare
National economic benefit or welfare can be measured by a variety of indicators. The most often used indicators are employment, preferably full-time equivalent employment, and gross domestic product (GDP). The two are used together because there will be times when employment will increase but productivity measured by GDP per person employed will fall. An unambiguous increase in national economic welfare will only occur if GDP and employment both increase and productivity does not fall.
The use of the GDP measure, especially in the case of mining expansion, is open to the criticism that it does not distinguish between foreign-owned product and product owned by domestic residents. The important consideration here is the distribution of gross product generated by overseas-owned enterprises. If 90 per cent of the gross product is distributed to domestic employees and in tax payments, foreign ownership is of little relevance. However, in mining approximately 60 per cent of value added accrues to foreigners in the form of interest payments, depreciation cash flow, dividends and retained earnings in the enterprise. These do not add to domestic incomes. An indicator which excludes foreign payments for interest, dividends and retained earnings is gross national product (GNP).
The other distinguishing feature of mining is its high level of capital intensity. Relatively large increments in investment are required to increase output. In turn, this means that there will need to be large deductions from GDP in the form of replacement investment appropriately financed from depreciation, if the output is to be sustained. A measure which deducts depreciation, or at least depreciation undertaken on behalf of domestic residents and foreign distributions out of value added, is net national product (NNP). This concept is akin to the concept of national net disposable income in the Australian National Accounts and measures the benefits to domestic residents in terms of household consumption expenditures and government finances.
Finally, it is argued that the prime measure of welfare is consumption expenditure. Hence, the relevant welfare indicator is the flow-on implications for household consumption expenditure plus net additional taxation receipts. Tax receipts are included because they determine further potential flow-on benefits for household consumption expenditure (tax rate reductions) or public consumption expenditure increases.
Gross benefits of the construction phase
The extent to which an episode of mining expansion benefits the prior-resident population of the country in which it takes place is strongly influenced by two factors:
- the extent to which mining investment in the construction phase increases demand and so increases employment; and
- the extent to which labour demands during the construction phase are met by immigration.
The first step in calculating the significance of the increase in demand is to assess the impact of mining investment under full additionality; that is, where the increase is met in a way which does not diminish other demands including other investment. This estimate can then be modified according to the extent to which the increase in mining investment reduces other investment.
The most practical and transparent analytical framework to estimate the demand effect under full additionality is to utilise a set of input–output tables that reflects the structure of the economy at the time the investment takes place. For this study, this was provided by a set of tables updated from the 2005– 2006 table to 2008–2009 with the distribution of value added reflecting the impact of foreign ownership. The algebra underlying the calculations is given at the end of the article.
From the 2008–2009 updated tables, the structure of the economy reflects the commodity price relativities which are likely to prevail, at least until 2014. Earlier input–output tables do not reflect current price relativities. If there are no capacity constraints on production, the increase in demand due to mining construction is measured by Type II multipliers; that is, the total of intermediate plus private consumption flow-on. These are given in Table 1, which is based on a flow of $A33 billion average annual net additional mining investment during the construction phase.
The modelling indicates that total imports grow by $A17 billion, which limits the increase in headline GDP to $A23.3 billion. If indirect taxes are added this gives a traditional multiplier of 0.74. If a Type III multiplier analysis had been employed, including induced non-mining investment flow-on, the multiplier would have been closer to unity.
A key variable of interest is the employment impact. The answer from Table 1 is that just under 200,000 additional workers will be required. The other key variable is the flow-on orders to manufacturing. From more detailed work, the total increase in gross output of the MM sector (from iron and steel to other machinery and equipment) comes to just under $A4 billion or a local content for the industry relative to total expenditure of 12 per cent.
These calculations provide an outside estimate of the value to the economy of the increase in demand due to mining expansion. The actual value is likely to be less, for two main reasons. First, the construction boom may directly attract labour and capital from other activities. Second, price effects associated with the boom may reduce the utilisation of both labour and capacity in other industries without transfer to the booming sector.
The first of these effects is expected and, indeed, welcomed by the neoclassical economists who dominate Australian policy formation; the latter, colloquially the ‘Dutch disease’ effect, they prefer to assume away.
Migration and mining expansion
Prima facie, there is little reason to expect major reductions in capacity utilisation in the non-resource industries as a result of the mining boom. On the capital side, much of the capital is supplied from overseas, both financially and physically. Indeed, the complaint is not that demand in the mining and construction industries is stretching capacity in the Australian capital goods industries but rather the reverse: that insufficient demand is flowing to these industries to offset the Dutch disease effect. On the labour side, the increase of less than 2 per cent in total labour demand is a lot less than the current unemployment rate, especially when Australia’s low labour-force participation rate (compared at least to some European countries) is taken into account.
However, it is no simple matter to shift underemployed labour into the jobs created by a mining boom. There are several reasons.
- Many of the jobs call for specialised skills. Here there is something of an impasse: Australian governments (Commonwealth and state) expect either that private individuals will see the opportunities and acquire the necessary skills, or that the mining and construction industries will provide the necessary training. For their part, and with occasional exceptions, the mining and construction companies expect that individuals and/or governments will ensure that the necessary skills are available. The result is that skill-specific labour shortages can easily occur.
- A significant minority of the jobs generated are located in remote areas where housing is poor but expensive and social opportunities are limited. Once again, there is an impasse: with honourable exceptions now mainly past, neither governments nor mining companies are willing to invest in remote mining towns which may lose their raison d’être within decades, and even when the physical infrastructure is provided the settlements often remain unattractive due to their isolation. The fly-in fly-out alternative gets round some of these difficulties at the cost of creating difficulties of its own, particularly for families.
- Mining and construction involve the operation of valuable equipment. The companies accordingly impose tight labour discipline and are quick to sack workers who breach discipline. Poor labour relations do nothing to assist recruitment.
The mining and construction companies attempt to overcome these disadvantages with high pay. They also call for immigration as a way of meeting their recruitment problems. If recruitment were purely on a guest-worker basis, this would parallel their attitude to the supply of capital and ensure that the mining expansion was essentially an offshore matter. However, Australia has no tradition of guest-worker migration, preferring permanent migration. This may be a realistic attitude in view of experience elsewhere (that guest workers become permanent) but may also reflect folk memories of the high transport costs once associated with migration to Australia and the resulting difficulties in attracting migrants. The result is that considerable migration is required to satisfy the limited and specialised labour demands of the boom, given that only a small minority among migrants is directly suited to fulfilling these demands. Therefore, we take a macroeconomic approach to assessing the immigration requirements of a mining expansion.
Given the strong employment impact of the construction phase, the expectation would be for net Australian migration to be correlated with mining investment. From Figure 1, typically Australia has responded to each episode of elevated mining expansion with increases in net immigration. As Figure 1 indicates, for the years 1980 to 1989 the level of net immigration averaged 105,000 a year compared to 60,000 for the 4 years before 1980. At the end of the 1990s there was another spike in net immigration, partly as a lagged response to the second construction phase.
Before the third mining construction episode the average was 150,000. However, from 2006 to 2010 the average was 260,000 a year, or a total cumulative increase in population of 550,000 compared to trends before the current episode of mining expansion.
If mi in the equation is set at zero and compared against the actual outcome, the estimate of the increase in population due to the mining boom falls to an additional 400,000 population between 2006 and 2010. This is lower than the 500,000 estimate given above from Figure 1 because of the influence of the trend term in the estimated equation.
If the average worker to immigrant population rate is between 0.3 and 0.4 (allowing for children and the basic male demands of the construction industry) immigration would have yielded approximately 150,000 workers towards the labour requirements of the construction phase. Therefore, between 60 and 80 per cent of the employment required by the current episode of mining expansion construction phase has been supplied by imported labour.
This dynamic is ignored in most assessments of the impact of elevated mining expansion. However, it is critical to the evaluation of economic benefits. The figure shows that a net 950,000 persons (rounded up to one million to allow for very modest net natural increase after migrant arrival) have been added to Australia because of the employment opportunities created by the three episodes of mining expansion since the late 1970s.
Significance of mining for the balance of payments
Where businesses are overseas-owned, employ little local labour and exist purely to supply overseas markets, a case can be made that they should be excluded from the national accounts of the host country and, instead, covered as offshore extensions of the owning economy. The case for this is particularly strong where labour is also supplied from overseas. Such enterprises were common in the 19th-century and early 20th-century empires and were often called enclave-export industries. Given that they frequently depended on franchises awarded by the colonial government, it is no surprise that with the end of colonialism they frequently found themselves nationalised.
The effect of excluding enclave-export enterprises from the national accounts of their host country would be to reduce the reported values for a number of variables, chiefly the following:
- the inflow of overseas investment, particularly during the construction phase of the enterprise;
- the stock of overseas investment;
- the outflow of funds to service overseas investment, both profits and capital repatriation; and
- the level of exports. (Sales by the excluded businesses would no longer be counted as export revenues of the host economy. Instead, resource rents collected by host-country resource owners (both governments and private owners) would be shown as exports, as well as labour supplied and any other sales by domestic businesses to the enclave-export industry.)
The net effect would be a major diminution in the importance of the excluded industries as measured by GDP or the national balance sheet, although as already remarked above there would be no effect in their importance as measured by NNP. The mining companies are strongly opposed to being defined as enclave-export industries, and equally to the use of measures such as NNP, because they wish to be regarded as pillars of the Australian economy. Their aversion to being regarded as overseas operators who just happen to be extracting resources in Australia is easy to understand given the history of nationalisation of mineral resources in countries where exploitation has been completely in overseas hands.
The decision to define mining operations as domestic rather than enclave-export industries has major effects on the headline economic indicators of host countries during periods of mining expansion. In particular:
- commentators focus on the growth rate of GDP rather than the relatively low growth rate of NNP;
- the growth rate of gross exports is emphasised at the expense of the lower growth rate of net export earnings after servicing overseas investment; and
- the terms of trade are calculated on the basis of the price of exported products rather than the price of services to the enclave-export industry.
In a rational world these effects would not matter much, but in the harried world of finance it is likely that they contribute to the Dutch disease. In what follows, the indirect effects of periods of mining expansion on resource allocation are considered.
As noted above, an episode of mining expansion requires a burst of capital investment. The benefit of the episode will be greatest if it uses otherwise unemployed resources, but it is more likely that resources will be transferred from less promising investments, or from less productive activities, into the construction effort required to expand mining. There is a double danger in this process.
- The price mechanisms which assist this transfer may themselves create offsetting unemployment. This can happen when they dampen demand for non-mining products and services without affecting the transfer of resources into the expanding sector. Instead these resources become unemployed.
- It produces a lingering shadow, in that investment foregone in the non-mining industries during the construction phase permanently affects their competitiveness. This mechanism was discussed in a previous article and its significance will be assessed below (Brain, 2012).
A particular case of these problems, referred to as the Dutch disease (because it was first recognised in the Netherlands during the North Sea oil boom) has two parts:
- During the construction phase the exchange rate is overvalued, in the sense that it renders industries uncompetitive when they would be competitive at the exchange rates which obtain both before and after the period of mining expansion. This reduces non-mining exports, a matter of some concern when mining commodity prices fall back towards pre-boom levels. Because investment in product development has been curtailed, the reduction in non-mining investment during a period of resource expansion carries the risk that production in non-mining trade-exposed industries will be reduced more or less permanently.
- The possibility of further episodes of overvaluation increases the riskiness of investment in non-commodity trade-exposed industries, so further reducing exports and import-competing capacity in these industries.
But why should the exchange rate be overvalued during the construction phase? In neoclassical theory the exchange rate cannot be overvalued or undervalued, because the foreign exchange markets are believed to take into account all relevant factors, focusing on the balance of trade and, hence, the ability of each debtor country to service its debts. Sufficient to say that during the era of fluctuating exchange rates (broadly since the mid-1980s) the Australian exchange rate has failed to behave as theoretically predicted. Reasons for the current overvaluation of the Australian dollar which relate to the decision to treat enclave-export industries as integral parts of the Australian economy include the following:
- an apparent high level of capital inflow;
- an apparent prospect of high growth in export revenues; and
- a high terms of trade.
In addition, other factors may be important, including the option exercised by a number of trade-surplus countries to maintain their exchange rates at ‘competitive’ levels by accumulating the financial assets of the indebted world (including Australia). It is also relevant that Australian real interest rates are higher than in the rest of the world, partly to restrain the level of economic activity (and, hence, assist in the transfer of resources to mining investment) but also because of the need to finance the high level of foreign debt.
For the purposes of the present study we now require a practical assessment of the severity of the Dutch disease as it currently affects Australia. The assessment requires comparison of a base case in which the mining industry continues its 1998–2005 growth rate through the period 2006 to 2012 and the actual case in which mining expansion occurred. The National Institute of Economic and Industry Review (NIEIR) has used its modelling system to make this comparison and published the results in the State of the Regions Report 2012–13 (Chapter 10). In brief, during the construction phase to date the mining expansion has, through mining investment and its multiplier effects, generated an increase of just under 7 per cent in Australian GDP. However, there have been offsetting reductions in activity in the non-mining industries which reduce the net benefit to date to approximately 3.1 per cent. This falls further, to 2.7 per cent, if the calculations are converted to a NNP basis. Taking into account the increase in population due to additional migration induced by the increase in mining activity, this further reduces to an increase of approximately 1 per cent in NNP per capita.
The State of the Regions Report further points out that all regions in Western Australia have unambiguously benefited from the mining boom. The regional pattern of benefit in the other states is patchy at best. The same goes for industry patterns, with strong increases of activity flowing from mining investment into construction while import-competing and non-mining export industries do less well than they would have in the absence of the boom. Among the industries which, on balance, do badly metal and machinery manufacturing is prominent. Given the importance of metal and machinery inputs to mining investment, this is an unexpected result, and raises the question as to why the linkage from mining investment to domestic demand for MM is so weak.
Local content of mining activity
Using NIEIR’s input–output table, the local content of mining activity for 2009 can be estimated. The direct demand from mining is defined as operational demands for goods and services plus investment requirements, including replacement investment. The operational demands are simply the intermediate industry demand column sums from the estimated input–output tables for all mining industries. This comes to a total demand for the mining sector of $A85.7 billion, of which $A75.8 billion is supplied by local industry. The bulk of the imports come from the metals and machinery (MM) sectors of overseas economies. Total MM demand for operations is $A11 billion, of which $A4.4 billion is supplied locally.
More importantly, from the investment data which is included in the totals in Table 2, it can be calculated that the investment component has a much higher import content. The import content is approximately half, with imports of $A26 billion. This includes replacement investment. Of the total imports, 60 per cent come in the form of products from the MM sector overseas.
These averages will change from year to year as the industry/project mix changes. For example, as the share of liquefied natural gas (LNG) projects in total mining investment increases over the next 2 years the MM sector’s share can be expected to increase.
Effects beyond the construction phase
These calculations place us in the position to assess the impact, not only of the construction phase but of the production phase which follows. The more the increase in mining capacity is bought at the expense of capacity in other industries, the less these industries will be able to respond to increases in demand during the production phase. We investigate these possibilities on the basis of the two extreme sets of assumptions (full additionality and full displacement) and also consider the consequences of the resource requirements to support the induced population increase.
Mining expansion: Full additionality
Full additionality occurs when impact of both the construction and production phases on the economy is close to the Type II multiplier estimates from the latest input–output table. This calculation shows what would happen if no capacity is lost in the non-mining industries during the construction phase.
Given the Australian Bureau of Agricultural and Resource Economics and Sciences (ABARES’s) benchmark for forecast mining production over the next 5 years, Table 3 shows the impact of each annual increment in production. ABARE’s projection implies an annual average increase of $A12.6 billion in gross mining output, or $A63 billion between 2010 and 2015.
From Table 3, the increase in GDP is $A12 billion, of which $A6.8 billion is mining gross product. Thus, the total mining gross product multiplier would be 1.76 under full additionality. Much of this is overseas-owned, and GNP increases by $A7 billion. What is important, however, is the increase in real NNP, or what the Australian Bureau of Statistics now calls real net national disposable income. This increases by $A8 billion, or 1.18 times the increase in mining gross product. The increase in real net national disposable income is two thirds of the GDP increase because depreciation and foreign transfers from the mining industry (which include repayment of foreign loans, interest, dividends and retained earnings) reduce mining NNP to 53 per cent of mining gross product.
Government revenues increase by $A2.8 billion while total employment increases by 57,200. This means that over the next 5 years a total of 286,500 employment positions will be created by the projected mining output expansion. If full additionality applies there is every reason to welcome a mining expansion.
Mining expansion: Full displacement
Gross full displacement is defined as the Type II multiplier that results when the increase in mining exports is neutralised completely by an increase in imports and a reduction in non-mining exports. This adjustment is modelled by applying the same percentage adjustment to all import penetration ratios and non-mining exports. In this case GDP falls by $A8.8 billion and NNP by $A7.2 billion. Employment falls by 78,100, reflecting the relatively high labour intensity of the sectors displaced. This is further reflected in the near $A5 billion fall in wages and salaries.
Net full displacement can be estimated by subtracting the second from the first column in Tables 4 and 5. Compared to the full additionality case, displacement considerably reduces the benefits of a mining boom, but in Australia’s case does not completely take them away. The numbers indicate that, except for employment, policy-makers can be indifferent to a mining expansion which results in full net displacement because, on this basis, there would be a key benefit of an increase in national productivity. Government revenue is also positive with net full displacement.
The issue of government revenue
A first caveat to this reasonably positive result is that the tax revenue for new mining projects may take 5 to 10 years to peak because of the high write-offs in the early stage of production for, for example, preliminary expenses, depreciation and exploration expenditure. Second, for highly capital-intensive projects, resource rent tax may not be levied for 8 to 10 years from the commencement of production. The taxation results in the tables assume immediate payment of resource rent tax based on the industry averages. For the full additionality case, the increase in direct tax revenue from the mining sector equals $A1 billion. In the early years at least the estimate of nearly $A3 billion government revenue from mining should be reduced significantly.
Costs of additional population
The second adjustment that must be made is that the gross and full displacement analysis indicates that 78,000 jobs are lost from industries that are, in the main, located in established regions with adequate infrastructure and community services.
The mining expansion case, in contrast, involves the necessity to increase national population in regions where, in the main, the infrastructure and resource requirements have not been provided. These costs need to be taken into account.
It is argued above that an additional 400,000 people will be induced to migrate into Australia to resource construction to allow mining production to increase as analysed in the tables. Given the rigours of life in many of the regions where mining investment is taking place, the workers imported to support the mining expansion will, by attrition, shift to industries and regions that are unrelated to the mining construction and production supply chain. If inflationary pressures are to be avoided, part of this attrition will have to be replaced by additional migration over the next few years. It will be assumed that the additional immigration will average out at 50,000 a year for the duration of the expansion, which is likely to continue to 2015 at least, and later for LNG if not for metallic minerals. Therefore, a notional 10 year construction phase will be assumed.
The annual and once-off expenditures that will be required over the next decade to support each 50,000 increment are given, by component, in Table 6 and come to a total of between $A5.0 billion and $A6.0 billion (to give a range rather than a point estimate). Given these resource claims from mining expansion either directly on government or in terms of resource claims on society as a whole (as is the case with housing), in order for Government and society to be indifferent between resource expansion and leaving the minerals in the ground the net Government revenue from resource expansion would have to be at least $A1.5 billion to $A2.0 billion greater on an annual basis than what is likely to be generated. If the balance is not to tip decisively away from net benefits to net costs there will have to be higher mining taxes, both in the short term and the long term and/or low levels of displacement of domestic non-mining production.
Three different tests can now be applied to assess the realism of these calculations. The first will examine the statistical relationship between real net national disposable income and mining product, the second will examine the relationship between state capacity and mining investment, and the third will examine the evidence of crowding out in the MM manufacturing sector. The three tests corroborate the reasoning in general, although they also suggest additional lines of investigation.
Link between mining and net national income
One simple way to test for the link between mining gross product and net national product/net national income is to run a regression of real net national income (less mining and construction gross product) per capita against time and real mining gross product. To remove the terms of trade effect, nominal mining gross product is deflated by the Australian National Accounts’ gross national income implicit deflator rather than by mining prices. The mining variable is expressed in per capita terms.
The zero coefficient for the mining gross product variable indicates that in both real and price terms mining has had no impact on non-mining and construction national net disposable income. At this level of abstraction, mining seems to add to income overall, although without any positive multiplier effects into other industries.
The results suggest a degree of additionality. In the full additionality model run considered above the ratio of the increase in NNP to the increase in mining gross product was 1.18. However, the equation just estimated indicates that full additionality does not apply and the mining production phase has contributed no more than mining’s direct contribution to NNP. The contribution from the full additionality sensitivity analysis is 0.53 per dollar of gross product, or an increase of $A3.6 billion a year between 2010 and 2015. Therefore, taking the ratio of $A3.6 billion to the $A8.0 billion increase in NNP from Table 3 suggests an average net additionality factor of 45 per cent, suggesting that while Australia has not avoided the Dutch disease it has at least managed to gain some net benefit from the construction phase. This case is referred to elsewhere in this study as the 50 per cent gross crowding-out case or the 50 cents in the dollar crowding-out case.
However, the conclusion does not adjust for the increase in population to support mining expansion. This is done in Table 7.
The conclusion from Table 7 is less optimistic than the conclusion derived in the section above on the Dutch disease, where it was found that the mining construction boom from 2006 to 2012 added approximately 1 per cent to NNP per capita. Table 7 relates to a longer time period which includes both periods of mining expansion and periods when the mining industry has been producing but not expanding. The conclusion for this longer period is that the annual mining contribution to net national income has been a little less than the per capita net national income attributed to the one million population increase to support mining construction phases since 1979. In other words, there has been no net additionality in terms of benefits to the original population, defined as the population that would have existed if net mining investment since 1979 had been zero. This means that, on a per capita basis, Australia has been subject to the resource curse, if not in terms of headline outcomes, and suggests that the main impact of the mining expansion since 1979 has been to expand population without loss of per capita net national income but no doubt at a cost of housing shortages and decline in infrastructure quality.
Resource expansion: Impact on non-primary GDP capacity
Our discussion in a previous article of the difference in investment drivers between commodity producers (now chiefly miners) and producers of differentiated and branded goods and services indicates that a major driver of displacement is foregone non-mining capacity that results from the pressures of the construction phase (Brain, 2012). Time-series estimates of non-agricultural capacity utilisation are available at the state level from National Australia Bank surveys on a quarterly basis and can be readily adjusted to exclude mining.
To test for the crowding out of non-mining activity during the construction phase the following model was estimated for the five main Australian states (estimation from the fourth quarter of 1989 to 2010):
Because of the interaction between import share and capacity it is not possible to interpret outcomes directly from equation coefficients. However, simulations suggest that for a $33 billion annual mining investment, the decline in MM gross output will be around $12 billion if the previous peak import share is exceeded or approximately $8 billion if this is not the case. Given the gross product to gross output ratio, this suggests a loss of output in terms of gross product of between $3 and $4.2 billion directly and perhaps double this after inter-industry flow-ons. This, in turn, would represent a plausible one third to one half of the reduction in aggregate displaced capacity estimated in the previous section.
Household debt and net additionality
We now have an apparent contradiction. It would appear that there has been headline net additionality, if not on an adjusted per capita basis. The results from the last two sections suggest a situation close to zero net additionality. The two can be reconciled by the recognition of the role of household debt in driving Australian economic growth since the early 1990s.
The accumulation of household debt stimulated growth through equity withdrawal by the household sector, including an increase in debt above the levels required to finance housing and other household investment. The net additional stimulus to growth is measured by the change in household equity withdrawal as a percentage of GDP. This series is given in Figure 2. Between 1992 and the middle of 2008 the average annual stimulus from the change in equity withdrawal was 0.5 percentage points, so that the growth in household debt contributed at least half a percentage point to annual growth. This was a powerful mechanism for producing headline positive net additionality which coincided with the construction phase of resource expansion – and was related to that phase by overseas willingness to lend to the Australian banks. The inference is that without the growth of household debt, mining expansion over the past two decades may well have produced headline zero net additionality. Under these circumstances it is likely that population growth would have been less. However, significant negative per capita additionality may well still have occurred.
The problem is that with the ratio of household debt to net disposable income now at 200 per cent there is only limited further stimulus to economic growth available from this source. It follows that it is likely that the economic stimulus from the current episode of elevated mining expansion over the next few years is will fall well short of expectations based on the last decade.
Macroeconomic policy and crowding out
An important link in all of these effects is the overvaluation of the exchange rate. We therefore revisit the question of why the exchange rate appreciated as it did.
One suggestion is that the appreciation could have been avoided by tighter economic policies. If, at the macroeconomic level, the objective is to protect the domestic non-resource sector an appropriate response to ensure internal/external balance would be to introduce contractionary fiscal policies to prevent the exchange rate appreciating and release sufficient labour resources to resource both mining construction and the maintenance of the non-mining sector activity at pre-mining boom levels. However, Australia took a less painful alternative, at least in the short-term, which was to increase the supply of labour by immigration so that both domestic supply and demand could be expanded and upward pressure on the current account balance neutralised. Australia imported 400,000 additional people which would have been more than enough to prevent crowding out of the non-resource sector and should have been more than enough to prevent upward appreciation of the currency. However, the estimated equations suggest that crowding out was not prevented. Despite a significant deterioration in the current account deficit (Figure 3) the exchange rate appreciated significantly between September 2004 and June 2008. The economic textbooks suggest that this should not have happened – the exchange rate should not have appreciated either due to the increase in the current account deficit or because of the additional population.
What then drives the exchange rate? The latest theory of financial sector analysis is that the high Australian exchange rate is being driven by the Australian currency being a ‘proxy’ for the Chinese currency. The Chinese currency is not fully convertible and is a non-market driven, controlled currency. The Australian currency is market driven. Therefore, international investors are reluctant to invest directly in Chinese financial assets to capture the benefits of China’s economic growth. They reason that the safest next best strategy is to invest in Australian dollar-denominated assets which, because of the dependency of the Australian economy on the Chinese economy, should in value ‘shadow’ the free market outcome for the Chinese economy. The inference is clear. Australia has no ability to control crowding out by macroeconomic policy instruments. The only solution is direct intervention to increase the direct benefits to the Australian economy from elevated periods of mining expansion.
This is not to suggest that increasing the migration rate has not lessened the degree of crowding out from what would otherwise have been the case. What is clear is that textbook policies are necessary, but not sufficient factors, to reduce the degree of crowding out. The textbook policies being insufficient, direct action is required if to crowding out direct action is required if crowding out is to be minimised.
Mining expansion and the national productivity slowdown
The central argument of this article is that the mining boom, by crowding out non-resource activity, has created unutilised resources which can be exploited to increase the direct benefits from the current mining expansion. This argument applies not only to capital and labour resources but also to the potential impact of mining expansion on the rate of productivity growth. This potential stems from the relationship between the rate of growth of productivity and the rate of growth of economic activity. For this to be correct, the evidence must suggest that the current productivity slowdown is related to the slowdown in the rate of growth in the economy. As is indicated below in this section, this is the case. From the December quarter 2011 National Accounts, the rate of growth of productivity measured by GDP per hour worked has fallen from between 0.5 to 0.8 per cent, depending on whether the September or December quarter of 2010 is selected compared to the corresponding quarter a year earlier. Figures 4 and 5 leave out the poor recent quarters and run to the June quarter 2010. Even so, productivity growth was still trending down. What is important here is not so much what the recent rate of growth of productivity has been, but the extent to which the slowdown in productivity growth was due to the changes in the pattern of economic activity.
Figures 4 and 5 indicate such a relationship. In Figure 4, in general the higher the rate of growth of GDP the higher the rate of growth of productivity. This finding is expressed in Figure 5 by a larger gap between the rate of growth of total hours worked and GDP. This gap is larger the higher the rate of growth of GDP.
The sample period is the first quarter of 1989 to the second quarter of 2010.
The two key coefficients are the sum of gdpg coefficients and the cute coefficient. The cute coefficient indicates that the lower the capacity utilisation rate in the economy the lower the (labour) productivity growth rate, no doubt in part due to the underutilisation of overhead hours. The sum of the gdpg coefficients is 0.68, indicating that a 1-per cent growth rate of gdpg is associated with an additional labour productivity growth rate of 0.32 per cent.
To demonstrate this impact of economic activity on GDP growth, the 2007 calendar year will be compared with the 2010 calendar year. The 2007 calendar year was the last year of sustained high productivity growth over all four quarters. The average GDP growth rate for 2007 over the four quarters was 4.6 per cent, compared to 2.7 for 2010. The difference in growth was 1.9 per cent, suggesting that the growth difference would explain a productivity growth decline of 0.6 per cent per annum.
However, productivity growth has also declined because of a fall in capacity utilisation rates between 2007 and 2010. The average non-farm capacity utilisation rate over 2007 was 83.8, compared to 81.6 for 2010. The lower capacity utilisation rate explains another 0.6-per cent decline in productivity. The average labour productivity growth rate in terms of hours worked in 2007 was 1.9 per cent, compared to 0.3 for 2010. Therefore, the slowdown in economic activity between 2007 and 2010 explains 1.2/1.6, or three-quarters of the decline in productivity.
The crowding-out effects of mining expansion would have contributed to the decline in national productivity over the last year. However, to date, the largest contribution to falling productivity would have come from the unwinding of the fiscal stimulus (Brain, 2010).
In the years ahead, however, the cumulative effects of the Dutch disease, if allowed to continue, can be expected to reduce national productivity growth rates from levels that would have been expected given longer-term historical trends.
The important point is that the crowding-out effects of mining expansion are likely to have a larger negative impact on national productivity growth compared to recent past crowdings-out. It follows that the impact of measures to increase the direct benefits from mining expansion will have a positive impact on national productivity and, therefore, will be mildly anti-inflationary, because they will enable existing employed resources to be used more effectively.
Given the changed circumstances that are likely to prevail over the next few years, extrapolation of past responses to mining expansion into the future suggests that there may be little headline net per capita additional benefit. This will come at a time when the resource claims to meet the infrastructure and service demands of the increased population induced by the current episode of mining expansion will be presented in full, creating very difficult political and economic constraints, and adding to those from climate change. This will be compounded by the national productivity growth remaining below historical trend levels other than occasionally, such as in 2011–2012.
The alternative is to change the way resource expansion is managed to maximise its net additionality. As discussed in a previous article, such management would include increased harvesting of resource rents and measures to increase the domestic content of mining investment.
ABS (Australian Bureau of Statistics), 2010, ‘Australian System of National Accounts, 2009–10’, ABS Cat. No. 5204.
Brain, P. J. ‘Australia and the Global Financial Crisis: A Highly Efficient Policy Response at the Cost of Locking in Structural Imbalances’, National Economic Review, 65, pp. 1–22.
Brain, P. J., 2012, ‘The Mining Boom in Context’,
National Economic Review, 67, pp. 1–18.