Why is the Western Australian Economy So Reliant on Mining?

Following our article on the decreasing diversity of the Western Australian Economy, Gareth Parker on WAtoday has a couple of good articles on the same topic. Firstly on the increasing reliance of the Western Australian economy on mining and then on the implications of how in such an economy you are very dependent on supply chains stretching across the world.

Recently world events have further shown how reliant countries around the world are on each other – which is great if countries are on friendly terms, but terrible if they’re not.

This post is an attempt to bring together some of our thoughts on why the Western Australian economy is so specialised. Obviously the real world is more complex than any model, but we hope the content below might be of some use for policy makers when they think about the State’s economic future.

Sadly, Mr Parker uses Chamber of Commerce and Industry data, rather than than from the excellent Practical Economics’ post. We don’t know but guess it’s sourced from a tweet by the Chamber’s Chief Economist Aaron Morey here.

One thing to note is that the Chamber obtains a share of mining in Western Australia’s Gross State Product (GSP) of 47%, while we obtain a figure of 49%. We don’t want to get hung up on the difference between the two estimates, as both get the right answer -that Western Australia is very reliant on mining. But in case you were wondering, here’s why there is a difference.

Both correctly use current price industry Total Factor Income (TFI) estimates^[1]From ABS Cat 5206.0 Table 6.. However, Practical Economics uses a slightly lower denominator in the sum of the TFI of all industries. Our calculation is:

$169,578 million/$348,532 million = 48.7%

The Chamber’s estimate uses total GSP from the industry TFI series, which also includes a couple of components which the ABS doesn’t break down to industry level and are presented as a sum-of-industry total – “Taxes Less Subsidies on Production and Imports” and a “Statistical Discrepancy”. The Chamber’s calculation is:

$169,578 million/$361,761 million = 46.9%

There is no right or wrong way to do this, but that’s why there is a slight difference.

If we had to guess, while the ABS isn’t clear on this, Taxes Less Subsidies on Production and Imports probably includes mineral royalties and so a large share of this category would be generated by mining.

Therefore mining’s true share of the Western Australian economy could be greater than 50%!

But that’s not the point of this post, which is: why is the Western Australian economy so reliant on mining?

There are a few reasons, and together they indicate that any diversification of the Western Australian economy is going to be a very difficult task.

Ricardo and Gregory

The first and perhaps most obvious reason why Western Australia is the concept of comparative advantage that we all learned about in first year economics, or was it high school, it’s been a long time since we did either.

Comparative advantage is an economy’s ability to produce a particular good or service at a lower opportunity cost than its trading partners. By opportunity cost of a good we mean the production of other goods that an economy must give up to dedicate resources (labour and capital) to produce the good in question.

We don’t intend to explain the whole thing here, as we’d expect anyone reading an economics blog would have seen it at some stage, but by comparative advantage we don’t mean absolute advantage or a lower cost, although that could be the case. Ricardo’s famous England/Portugal Wine/Cloth example showed that even though Portugal was more efficient at producing both goods, it had a comparative advantage in only wine.

An economy is better off, that is increases how much it can consume, if it specialises in a good or service in which it has a comparative advantage and trades with other economies to obtain other goods and services. In Ricardo’s example, England specialises in cloth and Portugal in wine, and the two countries trade to obtain the good they do not produce.

The novelty of Ricardo’s result is that an economy doesn’t need to have an absolute advantage (i.e. be the lowest cost producer of a good), but an economy can have an absolute and comparative advantage in producing a good (e.g. Portugal and wine).

Western Australia may have both an absolute and comparative advantage in its resources industry. There are few other regions with our resources endowment and we are very good at extracting these resources.

The main point is that, even if Western Australia had an absolute advantage in producing some manufactured product, the cost of diverting economic resources (labour and capital) away from mining to the manufactured good would likely be prohibitive.

The second but related factor at play here is the economics of a booming industry – more commonly known as the ‘Dutch Disease’ effect. More recently we’ve called this effect the ‘two-speed economy’.

First advanced by Australian Bob Gregory in 1976, but now more associated with Corden and Neary in 1982 who referred to the decline in the Dutch manufacturing sector after the discovery of the Groningen natural gas field in 1959.

We won’t go into the whole thing, but in the presence of a booming export sector, the Dutch Disease model predicts:

• a shift in resources (labour and capital) away from lagging export sectors (e.g. manufacturing to gas extraction). This is called the ‘resource movement effect’; and

• a boom in non-tradeable^[2]These are goods and services that cannot be imported or exported, so are produce exclusively domestically. Haircuts are often given as an example of such a service industry sales from the income generated by the booming sector. This is called the ‘spending effect’.

The mechanism by which the booming export sector crowds out the lagging export sector is an appreciation of the region’s real exchange rate, or ratio of non-traded good prices to traded goods prices. For a country, much of this effect can take place through an appreciation of its nominal exchange rate.

However, for a region such as Western Australia, which does not have its own currency^[3]Although the Australian dollar is often seen as a ‘commodity currency’ which can mirror the mining industry’s fortunes., much more of this work is done through changes in domestic prices (e.g. wages).

Additionally, the original Dutch Disease literature dealt with an expansion in quantities from the booming industry, rather than the price boom we’re experiencing in Western Australia. Nevertheless, we consider the model still applies.

So how do recent Western Australian economic conditions line up with our model? We will look at employment by industry as in the era of the COVID-related migration pause labour truly is the scarce factor of production and is indicative of expanding and contracting industries.

Obviously, the mining industry and in particular the iron ore sector, is booming.

As others have pointed out, the Western Australian labour market is extremely tight, with annual^[4]Annual average of 2021 over 2020, ABS Cat No 6202.0. employment growth of 4.6% compared with annual growth in the labour force less than 4% in the year to December 2021. Annual average labour force growth has frequently exceeded 10% in the past few years.

Consistent with the comparative advantage, Western Australia has a relatively high share of mining employment relative to the rest of Australia – 8.5% in 2021 versus 2.0% nationally.

Tick for comparative and probably absolute advantage.

The Dutch Disease hypothesis would predict a movement of labour into mining and that’s exactly what’s happened. Even with the COVID pause on immigration, Western Australian full and part-time employment rose by 56,000^[5]This is Annual Average Growth, i.e. the average over 2021 compared to the last pre-pandemic year of 2019 and is from ABS Cat No. 6291.0, which is the detailed industry employment series. The number … Continue reading people over 2021 compared to 2019, of which mining accounted for a little under one fifth (10,283).

Employment gains to mining are likely to be at the expense of other industries, but it is a little more complicated than that.

Chart 1: Change in Western Australia Employment 2021 versus 2019, Annual Average, Total Persons, Full and part-time Employment

The second-order Dutch Disease effects are interesting. The lagging industry isn’t manufacturing, as this industry tends to be closely associated with mining in Western Australia.

Because we don’t have our own exchange rate, the impact of mining tends to enter domestic costs – the Western Australian Consumer Price Index rose by 5.7% between December 2020 and December 2021, the highest rise since December 1990^[6]Excluding the artificial Goods and Services Tax implementation adjustment in 2001. Similarly, Average Weekly Earnings rose by 7.3% over the same period.

Construction probably competes most closely for labour with mining – skilled, competent blue-collar workers. Construction employment has been static over the past two years and, as shown in the table from our previous post, value added (Total Factor Income) also fell between 2020-21 over the last pre-pandemic year of 2018-19^[7]These are financial year figures, as opposed to the more up-to-date calendar-year employment figures presented here.

Total Factor Income Growth, 2020/21 over 2018/19, Western Australia

	Current Price	Constant Price
AgricultureForestryFishing	4.7%	2.2%
Mining	57.9%	4.3%
Manufacturing	10.7%	13.0%
ElecGasWater	-2.9%	-1.3%
Construction	4.0%	-1.8%
WholeTrade	21.2%	11.8%
RetailTrade	10.7%	5.8%
AccomFood	-5.3%	-3.0%
Transport	-2.8%	-10.5%
InfoMediaTelecomms	7.7%	3.6%
FinInsur	4.6%	3.1%
RentHiringReaEst	6.3%	2.0%
ProfScienceTech	8.3%	3.1%
AdminSupport	3.4%	-3.7%
PublicAdmin	9.1%	8.5%
EducationTrain	7.9%	2.8%
HealthSocial	10.7%	10.4%
ArtsRec	-0.7%	-3.2%
OtherServ	5.0%	7.1%
OwnDwel	1.5%	1.3%
Total Value Added	26.2%	3.7%

Demand for construction seems to be high, consistent with a strong Dutch Disease spending effect, but the industry just can’t keep up supply and prices are rising – nominal (current price) construction GVA increased by 4.0%, despite the volume of activity contracting by 1.8%.

The spending effect appears not to have benefitted retail employment, but retail value value added (GSP) did rise over the period.

Retail is also undergoing the ‘Amazonification’ effect, where a greater share of retail is going online and so the margin not captured be West Australian businesses. We also examine some specific COVID restriction effects below.

The real winners, both in terms of employment and value added, are the Public Administration, Education and Training and Health and Social Services industries.

These are all mainly public-sector industries.

There have been specific policy initiatives in at least health and education, but the more important impact is that the Western Australian Government has been running strong budget surpluses while expanding public sector expenditure.

The golden goose enabling this is royalties from the mining sector, notably from iron ore.

That is the spending effect in our case is through royalties and public sector expenditure, rather than private-sector expenditure.

COVID Restrictions

COVID restrictions have obviously played a major role in economies around the world contracting over the past two years.

However, we argue that such restrictions did not have a major impact in the unique case of Western Australia thanks to our strong mining sector.

Yes, we holidayed at home and spent money on alterations and additions to our housing, but employment and value-added in these industries did not grow much.

We argue that the COVID lockdowns, which were moderate in Western Australia compared with other regions around the world, and work-from-home arrangements merely accelerated trends that were already ongoing.

By that we mean 10 years’ worth of change has been packed into two.

For example, domestic retail trade was already being affected by online sales, but COVID lockdowns accelerated the trend. Australia wide, online retail sales jumped from just under $2.0 billion per month at the end of 2019 to over $4 billion per month in late 2021.

The technologies and practices allowing work-from-home did not start with COVID – it was the necessity of having remote teams in the US software industry as well as the deterioration of many US urban centers as places to live and work – but COVID made these arrangements mainstream.

While not Western Australia-wide impact, but more of an impact of a region inside the Western Australian region, these factors hollowed out the Perth Central Business District (CBD) and left many empty office building and shops.

When we were young, Perth CBD shopping had a real prestige associated with it. No more. This Perth we knew is dead and, like many cities around the world, it must find a new role in peoples’ lives or perish.

Associated with this has been a substantial decline in Perth public transport ridership to levels not seen since the opening of the Mandurah line in late 2007. This is likely due to less people working in the city, as well as reluctance to share air conditioning with masses of people in a sealed tin can in the middle of a pandemic.

We don’t know if this will last, perhaps rising petrol prices will bring the patrons back? However, we (i.e. the State) need to fully consider whether large, new, public transport projects are viable, given shifting large numbers of people from the suburbs to the CBD might be yesterday’s problem.

Krugman’s Increasing Returns Economic Geography

Perhaps the scariest model to look at Western Australia’s economic structure is Paul Krugman’s model of new economic geography.

Krugman gets rightly criticised for few very wrong predictions, but we have always found his economic geography model useful for explaining the Western Australian economy. He rightly received the The Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel in 2008 partly for his economic geography model.

The implications from the model are scary because the other models presented above are dependent on a strong mining sector or because of temporary COVID restrictions. This leaves open the possibility that if mining were to falter, or COVID restrictions eased, something else could rise up to take its place.

Krugman’s model however, only requires that industries have economies of scale (increasing returns) – that is a fixed cost and a constant or falling variable cost^[8]Fixed costs are those that do not vary with production, such as the capital cost of a factory. Variable costs do vary with production, such as labour or materials. as production increases – for ‘footloose’ industries to locate near their largest markets.

And we aren’t anybody’s largest market.

Formally economies of scale occurs when an industry’s average total cost of production (total costs divided by total output) falls with increasing production.

This can be as simple as being able to increase production, which enables a firm to allocate its fixed costs across a larger volume of output.

This result holds even if smaller regions have exactly the same costs as the larger region.^[9]As an aside, industries can chase low wages and costs for their locations, but we aren’t exactly low cost either.

Prior to Krugman, the traditional reason as to why industries cluster and cities form is agglomeration effects, where industries located together can use by-products and increase the available intermediate product and labour pools. Firms can draw from deep pools of relevantly skilled labour and workers can change jobs without changing their location of residence.

The Kwinana Industrial Area is often cited as such an example in Western Australia, called Industrial Symbiosis here, but such industry clusters are difficult to develop.

However, Krugman’s model does not rely on agglomeration effects. The best way to explain the economic geography model might be by way of example. Consider a large carmaker, perhaps centered in South Korea. It’s home market is its largest market and so it sets up its main factory there as this minimises transport costs and takes advantage of economies of scale because it can allocate its fixed costs over a large volume of production.

Car manufacturing is a non-location-specific industry, which can take place anywhere, while agriculture and mining are location-specific industries as they must take place where land or mineral resources are located.

It might set up in other markets if they are large enough. The USA is in fact the largest market for real-world Korean carmakers and so they have production based there.

Now consider a small market, such as Western Australia, or for that matter Australia. Our South Korean carmaker must decide whether to produce in Western Australia, where it will weight up whether to:

• produce in Korea, maximising its economies of scale and reducing its per-unit cost of production, but incurring the transport cost from South Korea to Perth; or
• setting up a factory in Perth, saving transport costs, but having a relatively high average production cost in Perth, due to a high fixed cost per car, and not fully utilising its economies of scale in South Korea.

Therefore, the greater an industries economies of scale and the lower transport costs, the greater incentive for the industry to centralise near its largest market.

The modern car industry relies of scale of production and economies of scale. Car factories located in relatively small markets, like Australia, let alone Western Australia, are now rare.

Of course, we did once have a Holden (Assembly) Factory in Mosman Park and a Ford Factory in North Fremantle. But those were different times.

And Mosman Park/North Fremantle was a different place – don’t forget the sugar refinery with the now million dollar views!

While we use the example of a foreign manufacturer deciding to set up in Western Australia, the same economics would apply to a manufacturing entrepreneur in Wester Australia deciding where to produce. Producing locally would likely place them at a disadvantage to their competitors.

Now for the frightening part – population flows.

As non-location-specific industry (manufacturing in Krugman’s model) moves away from the small region, so does population^[10]In response to higher real wages flowing from a lower cost of manufactured goods in the larger region..

This further reduces the market size of the small region and makes it less likely to produce manufactured goods locally, which further reduces population size. This keeps going until the small regions has only location-specific industries left.

Western Australia is now reliant on a small set of location-specific industries – mining and to a lesser extent but still importantly agriculture.

Things are good while those industries are good, but if they aren’t good then things get interesting.

Clearly, in our example, labour from Western Australia is unlikely to move to South Korea at any scale. However, population loss would be likely if mining declined, as has been experience in many cities in the USA when their main industry’s fortunes deteriorated.

If we weren’t strong in mining, we might not be strong in anything at all. And people might not hang around to see how that works out.

We’re reticent to call it dangerous to be so reliant on mining, as this industry has been so good to us for so long. As we’ve mentioned before, we’ve seen many government strategies trying to move us away from mining, but the industry has kept powering ahead.

We may well have 50-100 more years of good times ahead, maybe more. But we’re also a geopolitical conflict with perhaps an associated export embargo to certain trading partners away from disaster.

Where to from Here?

We’ll cover this in a future post but note that we are very small and, due to the success of our mining industry, very high cost. This means diversifying away from mining, and perhaps domesticating our supply chains so they are less vulnerable to disruption from international events, is very difficult.

We are reliant on mining and very vulnerable to any downturn in that industry.

However, a region doesn’t have to become Detroit if its major industry declines.

As long the quality of life in Western Australia is good enough for people to stay and start doing new things, rather than leaving for other regions, then things won’t be a complete disaster.

However, with or without a strong mining sector, it will be difficult to domesticate our supply chains. We’re just too small.

Paying the occasional mineral processing plant to locate here isn’t going to change that.

Thanks for reading. We hope you found this post interesting.

If you have any comments, suggestions or corrections to our analysis then please contact us. This isn’t our full-time job and we generate Practical Economics content when we can, so we might not reply immediately, but we will get back to you.

Alternatively sign up for our newsletter below for more great content.