It’s now difficult to avoid the conclusion that the surge in residential building that has powered the economy since the mining construction boom faded is on the wane — albeit, handily, commodity prices have picked up considerably recently, meaning higher national income.

Data on building approvals from October showed the fifth straight monthly fall in trend approvals, and the third straight month in seasonally adjusted terms. The value of building approvals also fell — the only positive was a tiny rise in non-residential building approval value.

For building approvals, the trend is the key figure. Because most approvals are in the hands of local councils, they can be extraordinarily lumpy and hostage to factors like how many holidays there are in a month. But all the signs in the recent data point downward.

A look behind the numbers, however, gives a more complex picture. Approvals in New South Wales — i.e. Sydney — are holding up, while Victoria/Melbourne is coming off the boil and Queensland has declined significantly; Perth, which has been falling for a while, is continuing to shrink.

And in Sydney, housing approvals are holding up, but more particularly non-housing dwelling approvals have stayed at the very high level they reached earlier this year. Non-house dwellings — let’s call them apartments, for the sake of convenience — reached over 4000 approvals a month in the middle of the year in NSW, for the first time ever, and while that number fell in October to just under 3700, that was still the seventh highest month on record.

In assessing how the residential construction boom will taper off, we need to understand that Sydney is a different market to everywhere else, because of its apartment construction. The last time as many houses as apartments were approved for construction in NSW was at the start of 2012. Since then, the ratio of houses to apartments has steadily declined. In October, the number of houses being approved was 64% of the number of apartments. In Victoria, the comparable figure is 130% — that is, significantly more houses than apartments are being constructed in Victoria. There was a period in early 2015 when the number of house approvals fell below apartment approvals in Victoria, but it was only temporary. The comparable figure in Queensland is 120% — although the house:apartment ratio has been around 90% for most of the last two years.

And NSW has long built more apartments than Victoria; if we go all the way back to the start of the 1990s, NSW was approving twice as many houses as apartments when Victoria was approving thirteen times as many and Queensland two and a half times.

In both Queensland and Victoria, housing construction has actually held up — it is apartment construction that has fallen. But in NSW, non-house dwellings keep on keeping on, fueled by Sydney’s relentless house price growth. That’s one reason why there’ll be no construction industry recession, at least in NSW: even if the hunger for apartments in Sydney declines — as you’d expect it would at some point — it’s at such a high level that it will continue to fuel significant building activity for a time even as it comes off. In other capitals, however, the decline in apartments is well underway, with houses again becoming dominant.

Westpac economist Matthew Hassan made a similar observation in response to the ABS October data, noting that “the construction cycle is now turning down” but that the drop in approvals would take some time to translate into lower construction activity. High rise apartments, for example, could take up to two-and-a-half years to complete, meaning the downturn won’t hit activity until “well into 2018”. JP Morgan economist Tom Kennedy said about the data “approval volumes are on track to decline sharply”, predicting that “the impulse from residential investment to real GDP should fade in 2017”.

The irony here is that the government has pushed through its heavily compromised ABCC bill for an industry now facing different challenges than dealing with strong wages growth and high demand for labour. And that will flow into the rest of the economy via (even) weaker wages growth and, perhaps higher unemployment.

And for those — like our friends at the OECD — who think the next interest rate rise will be up and in 2017, the RBA is unlikely to be too concerned about the impact of long-term low rates on the housing market — except on apartments in Sydney, if they persist at their current high levels.