Real and nominal GDP composition over time

I have been thinking about business cycles lately, and the degree to which I think Australia might be at risk of a recession over the next 12 months (about which I am still thinking - and will hopefully blog soonish).

Anyway, I decided to explore the extent to which different classes of expenditure contribute to GDP(E) over time. For the heck of it, I looked at GDP in both current prices (nGDP) as well as in chain volume measures (rGDP).

What is interesting to see is the extent to which the terms of trade changes since the early 1990s (among other things) have affected the long-term composition of the various elements of GDP(E).

Let's start with imports and exports.

What to make of this difference? I think it is telling me our volume of imports (as a proportion of the economy) has increased dramatically, but in nominal dollar terms, it is less changed over time. For completeness, we have a couple of charts on the terms of trade and the value of the Aussie.

Let's now turn to the headline expenditure measures in GDP(E).

Here it appears household activity as a proportion of nominal GDP(E) has declined; but the volume of produce purchased is unchanged.

On to capital formation (which I find a littler harder to explain).

And some totals ...

Seriously, for such a long time period (from September 1959 until now), the compositional change in nominal GDP is the one to look at (the red line in the above charts). The real GDP series (aqua in the above charts) seeks to hold too many things constant, and in the process introduces artifacts into the historical comparisons.