Apologies to any others, but if you’re not a dyed in the wool economics tragic read no further.
Back in uni days (over 30 years ago now) the old Phillips Curve theory was still being tossed around (although with modifications from the 1950’s original) and it’s always been something I’ve had niggling at the back of my mind in a very simplified form. My simplified form (in the short term) runs something like this;
If the level of unemployment falls close to, or below, NAIRU (which has generally been thought of as at about 5%, although these days may be lower) doesn’t it make sense that there would be some upward pressure on wage costs and therefore prices?
In addition, isn’t it reasonable to say that those wage pressures would take some time to end up being reflected in actual consumer inflation? (note I’m bypassing the whole idea of rational expectations).
I’ve therefore been a close watcher of unemployment levels and core inflation data with a 12 month lag (an arbitrary lag imposed to allow for that delayed effect on prices). When we plot this relationship for Australian unemployment data from June 1993 to June 2015 (latest data allowing for the 12 month lag in inflation) what we see is below.
A few things appear obvious straightaway:-
- The supposed inverse relationship of the classic Phillips Curve is obviously less than robust across all data!
- The relationship appears far more robust as we go below about 5.6% unemployment
- No matter what kind of trend line you try and fit to the data (and the graph has a number from linear and log to trinomial) the convergence at 5% unemployment and 3% core inflation is obvious.
I’m certainly not trying to suggest that we should be looking at this in any way as a predictor of future inflation; although over the past year unemployment has fallen from 6.1% to 5.8% and would therefore suggest at least the possibility of a move upwards in core inflation over the next 12 months. But it does at least give me some sense that there is a sound reason for the 3% upper inflation target range for the RBA (particularly if we buy into more modern Phillips Curve thinking that would suggest the long term curve is simply the vertical line at NAIRU).
Any other geeks out there got a thought? I’d be interested to hear.