Will the $155b data centre boom pay off for us all?
Commodity booms have been a huge part of Australia’s economic fabric and sometimes even its national identity. Think of the gold rushes that triggered mass migration in the 19th century, the post-World War II period when the wool trade was so important that it was said that “we ride on sheep’s back,” or the massive resource boom from China earlier this century.
But there’s another spending surge that’s been talking about the economic world lately: a boom in plans to build huge shacks full of computers.
Investment in data centers could rival past mining booms, and the government is keen to attract that money to Australia, albeit with conditions. Westpac estimates this is a $155 billion pipeline: a huge sum.
But when we look beyond the enormous numbers, how much benefit does this abundance really do to the economy? Don’t big booms often cause economic headaches as well as gains? How can we avoid these pitfalls this time?
When there’s a big spending spree, it inevitably helps the economy to some degree, but don’t be fooled by the big numbers on data centers. I don’t believe the economic benefits are as great as they seem. There are also lessons we can learn from past resource booms that gave the economy a short-term boost but also left us with long-term problems.
First of all, it is impossible to deny the scale of data center spending. HSBC said Australia’s data center investment was the fourth or sixth highest in the world, depending on how it was measured. Equipment spending by IT companies rose to a record level of $11.8 billion in the period through March.
Many data centers are being built, or at least planned, because tech companies believe we will need them to meet the growing demand for artificial intelligence. Australia is an attractive place to build them due to its space, potential to be powered by green energy, relative proximity to Asia and security environment.
So what does the data center boom mean for the broader economy?
Deputy Minister of Science, Technology and Digital Economy Dr. Andrew Chartlon argued this month that Australia was once again the “lucky country” with its data center boom, as well as resources.
On the positive side of the ledger, economist Charlton noted gains from business investment and employment, but admitted they weren’t all that good.
Most of the technology is purchased from abroad; Therefore, although investment in these facilities has some local benefit, its impact on gross domestic product is not as large as the headline investment figures.
There is also a short-term benefit in terms of employment. Building the centers will require labor, but Charlton also acknowledged that the data centers will employ “quite a small number of people” once they are up and running.
Even so, Charlton says the key economic benefit of data centers isn’t really about “what’s in the shed” but “what they enable everywhere else.” The government believes the real economic magic from data centers will come as more businesses find ways to use AI to be more productive, or start new businesses entirely.
This is where the claimed benefits become much more difficult to prove or disprove, because who really knows whether AI will deliver the productivity gains its supporters envision?
It’s also not clear that the presence of a large number of data centers in Australia makes it more likely that businesses here will find productivity-enhancing ways to use AI. But the government argues that in the digital economy, nations with their own computing power have the advantage of developing AI “according to their own laws, for their own purposes.”
Finally, Charlton argues that the data center boom could help accelerate the transition to green energy, insofar as power-hungry centers need to organize their own renewable energy.
What about the downsides?
To be fair, Charlton acknowledges much of this, including that data centers consume enormous amounts of electricity and water. If property isn’t managed, this hunger could cause consumers to pay more for electricity, derail green energy goals, or raise water bills. The government has said it expects new data centers to bring enough new green energy to meet electricity demands: this expectation needs to be translated into reality.
The local impact of data centers (including noise, visual impact and their placement near people’s homes or schools) is also a real cost to society that cannot be ignored.
There is also the classical economist question: What about the “opportunity cost”? What else can we do with resources instead of building warehouses full of computers?
If the boom is as big as many expect, the electricians and builders needed to build data centers will need to be moved away from other activities, such as building homes to address the housing shortage, adding to cost pressures.
These are legitimate concerns about the capacity of the economy. As we’ve seen in past mining booms, demand for skilled workers can translate into higher costs elsewhere.
Other economists have other reasons to be skeptical of the data center boom.
First, there is the fact that major AI firms are foreign-owned; This means that profits will largely flow overseas once the construction phase of the boom has passed. This is largely what is happening with the gas export boom.
Independent economist Saul Eslake, for example, is skeptical about whether governments can afford the tax revenues from past mining booms.
Miners must pay royalties for the resources they extract from the ground, no matter how creative their tax structure is.
He notes that some of the leading players, such as tech giants like Meta and Google, have a history of shifting profits to regions with lower taxation, resulting in them paying low corporate taxes compared to their billions of dollars in revenue.
Moreover, there is a possibility that the wave of AI-focused investment may not live up to the hype of being the massive game changer many expect.
We’re still at the beginning of the AI journey, but so far economic research has yet to show any major productivity returns, despite all the publicity from AI giants (some of which are about to hit the stock market).
As a result, the claim that data center abundance will automatically be some kind of boon for the economy is far from a slam-dunk argument.
This doesn’t mean we should close the door on data centers: Whether we like it or not, AI is booming.
But this means governments must impose strict rules on how data centers affect electricity and water supplies and their impact on local communities. We must be aware that big numbers on data center investments are not the whole economic story.
Maybe Charlton’s optimism will prove true and this time we will become the “lucky country” again, thanks to the dizzying increase in artificial intelligence spending.
But that actually depends on AI playing a game-changing role in productivity, and that’s a question that likely won’t be answered for years.
The Business Briefing newsletter delivers big stories, exclusive news and expert insights. Sign up to receive it every weekday morning.
Clancy Yeates is deputy business editor. He covered banking and financial services and was previously a national business correspondent in the Canberra bureau.Connect with: X or email.
