Australia is building a lot of new homes. But is the datacentre boom holding us back? | Greg Jericho

The latest series of data on housing and building activity has been released, showing residential building is growing strongly but also that the datacentre boom is massive and ongoing.

As political parties look to keep housing on the agenda, the massive resources devoted to building datacentres is going to hit anyone wishing to increase housing supply.

For a government wanting a good story, it has one in the form of housing. Not only are house prices finally showing some semblance of reality, housing construction is also doing well. In the year to June, the volume of residential building rose 8.7% – the strongest since 2016.

The increase of total building was even stronger – up 10% on the year, the best since the GFC stimulus years:

Click here if this graph does not display

The government would also be extremely satisfied to see the number of new private sector house commencements rose 11.6% to 31,707 dwellings.

That number is the highest recorded since 1994, excluding the boost during the Covid “homebuilder” surge and the abnormal number of commencements in March 2000 designed to beat the introduction of the GST.

Click here if this graph does not display

The government will suggest this is a sign that investors are shifting to new housing construction, given that still gives them the 50% capital gains tax discount.

As the June quarter includes time before and after the budget, we will need to wait for the September quarter to get a better sense of whether or not this is a blip or a more sustained increase.

But there is decidedly no blip for public sector housing, which has – for over 30 years now – been barely part of the housing equation.

Click here if this graph does not display

This was a key aspect of a report released last week by the Senate’s select committee on intergenerational housing inequality, Locked Out: The Growing Generational Housing Divide.

The committee, chaired by the Greens senator Barbara Pocock and at which I gave evidence, made numerous findings of a pretty general nature. The first was unanimous across all members of the multiparty committee: “It is an indisputable fact, universally agreed by the committee, that Australia is in a housing crisis. This crisis has been long in the making.”

The report highlights that while the changes to the CGT discount and negative gearing were vital to improve the housing crisis, there remains still much to be done.

Public housing is a key part of that. The committee found “the dominant message from submitters [including myself] was the need to build significantly more affordable and social housing at scale through long-term government investment and clear delivery targets”.

While the committee itself made no specific recommendations, due to its multiparty nature, Senator Pocock proposed “that the Australian government commit to a national target to increase public and community housing to at least 10% of Australia’s housing stock over the next decade”.

This would mark a significant shift.

From 1955 to 1985, public and social housing accounted for, on average, 13% of all new residential buildings. By contrast, over the past 30 years (aside from the GFC stimulus period), public and social housing has never accounted for more than 4%.

In the year to June, it was just 2.3%:

Click here if this graph does not display

The report also noted the issue of rents, and Pocock also proposed a cap on annual rent increases at the lesser of 2% or the inflation rate.

The issue of rents is rather hot at the moment as the opposition tries to blame rents going up on the changes to the CGT and negative gearing (which is laughable, given anyone renting a place before May is grandfathered), and One Nation are claiming their cuts to migration will deliver falls in rents.

In reality, rents essentially follow interest rates:

Click here if this graph does not display

But all housing policy around new builds always needs to deal with the issue of the available number of builders to actually do the work. And this is where the datacentre boom is truly becoming an issue, because that is accounting for stonks of work.

Datacentres are covered in the Australian Bureau of Statistics survey under the “commercial building not elsewhere classified (n.e.c)” category. Usually that makes up about 1% of the value of non-residential building commencements. But in the year to June, it was 20%:

Click here if this graph does not display

So big has the surge of datacentre building been that it was worth just under three-quarters of the increase in non-residential building commencements over the past year.

We are putting a heck of a lot of eggs into the datacentre basket, hoping like hell it delivers benefits to someone other than AI companies selling hopes and dreams.

Usually almost double is spent building factories compared with the buildings in the “commercial n.e.c” category. But in the past year, more than 17 times the amount of money was spent building datacentres than factories:

Click here if this graph does not display

That’s a lot of resources devoted to something that, unlike a factory, will not employ many people after it is built, and unlike a house will not provide a place for people to live for the next 40 years.

Right now building construction is doing well. But the question is whether we will look back and see a waste of money and work on buildings that didn’t deliver what was promised, instead of money spent building homes for people to live in or places for people to work.

• Greg Jericho is a Guardian columnist and chief economist at the Australia Institute