From debt ratios to interest rates to unemployment, simple averages of the past are often not a good guide to where things are headed.
One of the frustrations of the post-GFC period was the way some people (and international agencies) assumed that historical averages of key ratios like housing prices or debt to income defined ‘normal’. If one of these macro ratios was well away from that historical average, it was an ‘imbalance’ that needed to be corrected, it was claimed.
The problem with this idea is that often the metric in question does not have a ‘right’ level that prevails through time. In the case of the economy-wide ratio of household debt to income, the sustainable ratio is higher in recent decades than it was back in the 1970s and 1980s. If inflation – and so nominal interest rates – falls permanently, the sustainable debt-to-income ratio rises, because households can service a larger loan with the same repayment. Financial deregulation also removed other artificial constraints on borrowing that prevailed back then.
This point has been well understood for more than 20 years, having been written about by various RBA staff members (including me) all those years ago. Yet still one hears concerned comments that once upon a time you could only borrow four times income, and now you can borrow a much higher multiple. And it’s true, because once upon a time inflation averaged 6–8% and mortgage rates were double-digit, but not anymore.
The misunderstanding was even more frustrating because, often, the historical averages used were based on data sets that went back to 1980. Since Australia was later to join the low-inflation club than many of its peers, more of the period since 1980 was in that high-inflation-low-debt era. That drags the historical average lower, making the recent data look higher in comparison than for other countries that already had inflation down by the early 1980s. That Australia looks ‘worse’ on these metrics is mainly a statistical artefact.
There is a broader point here: historical averages do not always represent centres of gravity to which the world must somehow return. Many of the metrics in question are emergent properties of the economic system and not bound to return to a particular number. We have made this point before, regarding the structure of interest rates globally and the sustainable level of the unemployment rate.
Part of the issue is that even if people behave similarly to the past, the macro-level averages and ratios that come out of that behaviour might not be the same as in the past. The composition of the population might have changed, or some other factor that changes the macro-level outcome. Certainly, the age structure of the population has changed. Population growth rates also do not stand still; in Australia, population growth has been noticeably faster post-GFC than pre-GFC. This has implications not only for labour market variables, but also things like the required rate of home-building each year.
Things aren’t the same after a shock
The question of where ‘normal’ is becomes especially salient when you are coming out of a large shock like a pandemic. It is tempting to look at the pre-pandemic period as the benchmark for where things are likely to return, but this is probably a mistake.
The reality is that the pre-pandemic period wasn’t ‘normal’ either. There was considerable labour market slack in Australia at the time. Wages growth consistently undershot RBA and other forecasts. Inflation lagged below target despite what appeared to be very expansionary monetary policy.
There was something going on beyond the national level, too. Many peer economies were finding that unemployment rates could decline to levels not seen in decades without wages growth or inflation picking up materially. Global rates and risk spreads were also far from normal, compressed to extreme levels. If someone had told me at the beginning of my career that large parts of the European corporate bond universe would have negative nominal yields for a sustained period, I would never have believed them.
Another decidedly non-normal feature of the period between the GFC and the pandemic was that business investment in many advanced economies (including Australia) lagged historical averages. So did trend productivity growth. These trends were probably related, with some researchers hypothesising that this was a consequence of the financial crisis, and the associated weak demand and debt overhangs.
The upshot is that the global economy had probably barely completed the adjustment to the previous big shock, the GFC, before being hit by the next one, the pandemic.
Make the trend your friend
How can you forecast, or even interpret current events, when the ground is shifting in this way?
One approach is to focus on the underlying behaviour at a more micro level and let the implications for macro variables flow from that. For example, forecasts of consumption are typically based on past experience of people’s spending responses to additional income. This approach won’t always predict actual outcomes: as Westpac Economics colleague, Economist Jameson Coombes reported yesterday, the recent data from the Westpac–DataX Consumer Panel is pointing to a smaller spending response to the Stage 3 tax cuts than the historically typical response. But it is better than playing chartist with macroeconomic ratios by assuming that consumption reverts to a ‘normal’ share of income.
It is also useful to factor in any longer-term trends that are in evidence. The trends in the labour market are a case in point. In addition to the stronger average population growth, the participation rate has been trending up for decades and this shows no signs of ending.
If population growth is stronger than it used to be in decades past, then employment growth needs to be higher to keep pace, too. And if the participation rate is trending up, employment growth needs to outpace working-age population growth to avoid rising unemployment. Some observers have interpreted recent rapid growth in employment as a sign that the labour market is still strong. But it could equally be viewed as being insufficient to keep pace with the even faster growth in labour supply.
It all depends on what your view of normal is.