Evidence for the long-term: economic growth could increase until 2026 then slow thereafter

If Brexit and Donald Trump have taught us anything, the world is getting tired of the slow growth status-quo. According to exit-polls from the USA election, the biggest reason for voter dissatisfaction is the state of the economy.

While it is not exactly clear what voters expect to see from the economy (unemployment is near cyclical lows and both economies have been recession free for several years), what is clear is that people are desperate for greater wage growth and GDP growth at a household level.

Therefore, we are digging under the hood of this economic concern by looking deeper into the population statistics. What we find is not particularly favourable in absolute terms, although we find that Australia and India are better placed than most. As a part of this analysis, we conclude Donald Trump is pushing uphill to “double the growth rate” as he suggests.

Is there a link between the population and economy?

The real economy relies heavily on population growth; both logically and practically. There are many ways we can verify this importance, however one of the best ways is to illustrate it over the very long-term by showing the 20-year trend of the economy, household spending and population growth. This is an effective method because it massages out economic cycles and shows that all three are intricately linked and generally mean-reverting in the long-term. Below is the case for Australia, showing a strongly aligned relationship.

The unemployment rate will be a critical factor to household spending in the short-term, however is generally seen to be cyclical. Therefore, while an unemployment spike is very likely to coincide with an economic downturn, over the very long-term it gets largely massaged out as excess capacity shifts in waves. Similarly, household savings could have a material impact in the medium term, as people spend more or save more depending on the state of the economy. However, in the very long-term this can also be expected to remain somewhat cyclical.

Employment considerations should be contemplated more broadly though. With an ageing population and extended life expectancies, we can anticipate that the percentage of the elderly workers will increase (as in the case of Japan). We may also see a higher percentage of individuals enter the workforce (especially in countries with a low female participation rate), which could provide further upside. These factors all materially impact the growth outlook.

What does it mean going forward? Why slow-growth may be the new-normal after 2026.

Using population and peak spender logic, we can illustrate the “population effect” and the added “ageing population effect” as per the chart below. This uses the Australian Bureau of Statistics expected population projection (accounting for expected changes in birth rates, death rates and an estimate of net migration) and is adjusted for the “peak spender theory” in the dark blue dotted line. We find the ageing population is expected to support the Australian economy in the short-term but will slow thereafter.

Take it as a grain of salt

While our analysis is a brave attempt to provide clarity and logic to an uncertain dynamic, it should not be considered a guarantee by any means. It must be understood there is significant scope for error. To provide full transparency on the scope for error as we see it, below are some of the points worth considering.

First and foremost, the entire premise of this analysis is that the economy and household spending are intricately linked. If this breaks down, the accuracy of the analysis breaks down. While we are confident in the link remaining to some extent (there appears no logical reason it wouldn’t), the reality is that household spending accounts for 55% of GDP which leaves 45% unaccounted for. Therefore, if we see a period of structural change in the remaining 45%, GDP could move markedly from the population trend.

There are also significant risks to any changes in what we’ve labelled as “household effectiveness”. The pace of innovation and technological development continues to drive progress at a household level, but our assumption this will continue at approximately 2% has inherent scope for error. Whether it is 1%, 2% or 5% is anyone’s guess in a world of robots and driverless cars.

The projection data is also subject to potential error, as birth rates or migration rates could change between now and 2100. There are also event risks, such as a world war that could materially impact both population numbers and household spending. A major asset market crash could have a similar impact and interrupt the trend significantly to the downside.

Lastly, from a technical perspective we have considered GDP in real terms throughout our analysis, meaning inflation has been discounted. The reason is that the relationship between population change and inflation is complex, due to the interaction between money supply and the velocity of money. Withstanding this, any period of hyper-inflationary or deflationary conditions could create further inaccuracies.