RDP 2013-02: Industry Dimensions of the Resource Boom: An Input-Output Analysis 1. Introduction

The Australian economy has been going through a period of structural adjustment in response to the boom in the terms of trade starting in the mid 2000s and the associated appreciation of the exchange rate (Figure 1). These significant changes in relative prices, and the associated boom in resource investment, have had very different implications for growth in the resource and non-resource parts of the economy.^[1]

In order for Australia to take advantage of the rapid increase in demand for its natural resources, not only has the resource extraction sector been required to grow rapidly, but so too have industries that provide inputs to resource extraction and investment. This ‘resource-related’ activity tends to be more intensive in the use of labour than the resource extraction sector.

The purpose of this paper is to estimate the size, growth rate, and industry composition of a broader measure of the resource economy, which includes both resource extraction and resource-related activity. This broader measure has been discussed in recent Reserve Bank of Australia (RBA) analysis including Lowe (2012), Plumb et al (2012) and Kent (2013). The main contribution of our paper is to outline the methodology, assumptions and limitations of this measure.

There are a number of reasons why it is important to quantify the linkages from demand for Australia's natural resources to activity in other domestic industries. Most importantly, it can provide an insight into the nature of structural change that is taking place in the economy as a result of the resource boom. Specifically, our methodology allows us to assess which industries have benefitted the most from the resource boom and, as a corollary, which industries will be most affected by any decline in the terms of trade and resource investment. Related to this, under certain productivity assumptions, we can also estimate the amount of labour that is required by each industry to service the demand for Australia's natural resources.

As outlined in Plumb et al (2012), it is useful conceptually to divide economic activity into three parts: (i) resource extraction, (ii) resource-related activity, and (iii) non-resource activity. We define the ‘resource economy’ to be the sum of resource extraction and resource-related activity:

• Resource extraction includes mineral and gas extraction, and also resource-specific manufacturing (such as the production of metals and refined petroleum). This is very close to the Australian Bureau of Statistics' (ABS) definition of the mining industry, the only difference being that we include resource-specific manufacturing in our measure.
• Resource-related activity includes investment that supports the future production of resources as well as the provision of intermediate inputs that are used in the current production of resources. In other words, it captures activities that are directly connected to resource extraction, such as constructing mines and associated infrastructure, and transporting inputs to, and taking extracted resources away from, mines. It also captures some activities less obviously connected to resource extraction, such as engineering and other professional services (legal and accounting work, for example).
• Non-resource activity includes everything else in the economy that does not have a direct relationship to the current and future production of resources. That is not to say that these other parts of the economy are not affected by the resource boom. Among other things, there are income effects associated with dividend payments to households, the benefits of tax revenue from resource extraction and resource-related activities, and spending by those working in those industries. However, only production, not income linkages, are considered in this paper.

In previous work, the RBA (2011a) has presented a demand-side measure of the resource economy, summing together resource exports and resource investment, and subtracting the imported component of that investment. While this is a simple and transparent way to measure the resource economy, it does not capture the entire resource economy (it excludes domestic final demand for resource output and does not adjust for imports of resource commodities, such as oil and petroleum), and it cannot be used to quantify the linkages that the resource sector has with other industries.

To address these issues, we use input-output tables from the ABS to transform a more comprehensive demand-side, expenditure-based measure of the resource economy into a supply-side, value-added-based measure that can be decomposed by industry. Input-Output tables allow us to identify the industries that provide intermediate inputs to resource extraction and investment, and to answer questions such as: to export $1 of iron ore, which industries would need to provide intermediate inputs for this $1 of iron ore to be produced, and how much gross value added (GVA) and employment would be generated by these industries as a result?

Our methodology builds on that in Kouparitsas (2011), which condenses input-output tables into three industries – agriculture, mining and ‘non-agriculture, non-mining’ – to examine the spillovers from mining production and investment to other industries.^[2] The contributions of our research are four-fold. First, we extend the methodology in Kouparitsas by using a finer level of industry disaggregation: we disaggregate the mining sector into 9 sub-industries, and the rest of the economy into 13 industries that align closely with ABS industry definitions. This allows us to estimate more precisely which industries have benefitted the most from the resource boom. Second, we include all final demand for the output of the resource extraction sector in our measure of the resource economy, not only resource exports. Third, we decompose resource investment by type of investment and allocate this to the industries responsible for undertaking the investment, which again allows us to estimate more precisely the industries that have been particularly important in the recent resource boom. Finally, the additional level of industry disaggregation used in our paper allows us to derive an estimate of the labour required to service the demand for Australia's natural resources.

In summary, our approach requires us to:

• First, estimate all of the final demand in the economy that is related to resource extraction and investment, and then identify the industries that produce these final goods and services. The industry that produces a final good or service is the industry that is responsible for the final steps in the production chain for that product. For example, resource exports are produced by the resource extraction industry, and resource-related construction investment is assumed to be undertaken by the heavy & civil engineering construction industry.
• Second, using input-output tables, we calculate the value and industry composition of intermediate inputs required to meet this final demand. For example, we calculate the value and industry composition of intermediate inputs that are required by the resource extraction sector to produce each $1 of resource exports, and the value and industry composition of intermediate inputs that are required for each $1 of resource-related construction investment undertaken by the heavy & civil engineering construction industry.
• After making some simplifying assumptions, we can then use the information in the second step to transform the final demand that is related to resource extraction and investment into estimates of GVA by industry.

The structure of the rest of our paper is as follows. The next section outlines our methodology, including a brief overview of input-output tables and how they are applied to our research question. Section 3 outlines our baseline results, and those for a narrower definition of the resource economy. The assumptions that underpin our approach are discussed and tested in Section 4. Section 5 considers the implications of strong demand for Australia's natural resources on employment by industry, and Section 6 concludes.

Footnotes

There have been a number of speeches and papers published by the RBA in recent years on the causes of the resource boom and the implications for the Australian economy. For example, see Plumb, Kent and Bishop (2012), Lowe (2012), Connolly and Orsmond (2011), RBA (2011a, 2011b), Stevens (2011b), Battellino (2010) and Connolly and Orsmond (2009). [1]

The methodology in Kouparitsas (2011) is implemented in Gruen (2011) and also summarised in the appendix to Gruen. [2]