Ripple effects from the pandemic are not over and underlying trends are not fixed. It would be a mistake to assume otherwise.
Whenever one reads commentary about what policymakers should do, rather than what they will do, one should always consider the author’s unstated assumptions. Two such assumptions seem especially pervasive currently.
The first is the belief that pandemic-related effects have already washed through and are no longer affecting outcomes. Increased working from home is usually acknowledged as an exception.
The second assumption is that the post-pandemic economy will fluctuate around the same average levels of key metrics as prevailed before the pandemic.
Neither can be assumed. The first is demonstrably false and the second cannot be justified by the evidence.
Statements that rest on these unstated assumptions include the idea that most of Australia’s current high inflation is demand driven. According to this view, monetary policy is still not tight enough and unemployment is too low. Yet the ripple effects of supply constraints unwinding are still evident. Attributing all, or even most, of the current high inflation to strong demand ignores this.
It also ignores the signs that growth in domestic demand was soft in the second half of 2023, particularly in the household sector. As Westpac Economics Senior Economists Andrew Hanlan and Jarek Kowcza flagged this week and the Treasurer has also indicated, we expect the December quarter GDP result to be soft – most likely flat in the quarter. The momentum in domestic demand is only barely positive.
Domestically, these ripple effects include the surge in population early last year, as Australia and China reopened borders. Because so much migration to Australia is by students, population flows are seasonal, spiking at the start of each university semester. This makes it hard to know the underlying trend in real time. We will only see this post-pandemic effect unwinding in the data over the course of this year. Rent inflation should ease in line with this, but it will take a while.
Related to this, the decline in average household size in Australia’s largest cities is also yet to unwind fully. Some of the decline was probably a response to lower rents during the height of the pandemic, as well as lockdowns and a desire for more living space when other options were curtailed. Some, however, could be a more lasting response to increased working from home. Because it is more pleasant to spread out than to crowd back in again, the reversal of the temporary part of the response is likely to be more drawn out than the original shift.
Together with the population surge, smaller households mean upward pressure on rents and housing prices for at least a while yet. But we can be confident that at least some of that pressure will unwind over time, regardless of what policy does. Historical norms are therefore not yet the right benchmark to use when assessing the current stance of policy.
The lingering ripple effects of the pandemic are not limited to the demand for housing. The supply side of this sector is also still constrained. The result is a backlog of partly built homes that is being worked down only slowly. The pandemic-era HomeBuilder program contributed to this by bringing forward some demand. Supply capacity in the industry is also constricted. This is partly because non-residential construction activity is competing for many of the same resources. In addition, bankruptcies and other exits from the industry have disrupted the networks of suppliers and subcontractors. These networks take a while to stitch themselves back together. In this way, temporary shocks can have lingering effects.
Pandemic ripple effects go beyond the housing market. Supply chains may have largely healed globally, but in some areas domestically things are still far from normal. For example, industry sources suggest that wait times for delivery of some makes of car are still well above pre-pandemic levels. Multiple disruptions at domestic ports have also degraded supply chain functioning.
The faults in our stars
Even once the ripple effects from the pandemic have faded completely, there is no guarantee that economic behaviour reverts to previous norms. Past experience is not worthless, but it would be a mistake to assume a static world where key economic metrics and relationships never change. Identifying shifts in these relationships – especially changes in trends – is one of the key challenges of forecasting.
Some of these trend shifts are more or less permanent, such as the invention and adoption of new technologies, the adoption of inflation targets at low rates, or financial deregulation. Step changes in one factor can also instigate ‘waves’: drawn-out responses in other factors. The step change to low inflation targets and financial deregulation induced a permanent increase in the sustainable ratio of household debt to income in Australia and other industrialised economies. This transition to higher actual debt took more than a decade to complete. The green transition will likewise induce a wave of investment for years.
Integration with the European Union induced rapid increases in income in Ireland and Spain, and so a wave of home-building to replace existing housing stock with higher-quality homes more suitable to current income levels. China is coming to the end of a similar secular ‘wave’ of high demand for new construction, in addition to the authorities’ policy actions to slow this sector. I have never seen that transition to a smaller construction sector end well, but perhaps the Chinese authorities can ensure theirs is the exception.
And some trends are a continuous evolution not a one-off shift. One example of this kind of wave is the increase in longevity, which has no obvious end point although it could revert under certain conditions.
So it is with the structure of interest rates and the feasible level of unemployment. These are not external factors handed down from the heavens: they are outcomes of the system. They evolve continuously.
There are reasons to believe that the structure of real and nominal interest rates will be a bit higher in the period ahead than it was in the years leading up to the pandemic. Shifting fiscal policy stances matter here, along with risk appetite (in part influenced by regulation) and technological change.
There are also reasons to believe that periods of low unemployment drag the feasible rate of unemployment down, as both new workers and employers gain experience in working with each other during tight labour markets. The RBA seems to understand this, even though the standard models they use are not well-placed to capture this particular dynamic.
The key point here is that these underlying trends – the ‘star’ variables – are not fixed. Both policymakers and policy-watchers would do well to continually test their views about these factors against the broadest possible range of data. Forecasters who do not allow for these potential changes risk being blindsided by them.
