The very first post on this blog was about identifying recessions, based on two consecutive quarters of negative GDP growth. At the time I noted the other definition of a recession: an 1.5 percentage point rise in the unemployment rate over a twelve month period.
Recently I stitched together the two ABS data series on the unemployment rate. The first is the quarterly series from August 1966 to November 1977. The second is the monthly series from February 1978. Because the quarterly series is original data, I have seasonally adjusted it so that it can be compared with the later seasonally adjusted data. The key chart follows:
If I overlay my original assessment of downturns (based on GDP data), you can see my problem. The recessions in the 1970s just do not look right. Also, the GFC qualifies as a recession on the unemployment rate, but just a slowdown on GDP data.
So let's try and mark unemployment recessions from the trough through to the peak, within which the unemployment rate increased by at least 1.5 percentage points within a year. Note, I also count the late 1970s, which only made it to 1.4 percentage points of unemployment growth on trend in a 12 month period.
Just for the fun of it, I can mark both sets of highlights on the same graph. This is a little ugly, but allows for a comparison of my periods. The unemployment method is highlighted in the upper half of the chart and the GDP method in the lower half of the chart.
Conclusions: there is a reasonable degree of agreement between the two indicators of a recession (two quarters of negative GDP growth versus a 1.5 percentage point increase in the unemployment rate in a year). However, although the GDP measure suggests a recession in the early 1970s, it is more accurately described as a down turn. Furthermore, the GFC did create a short-lived recession in Australia.