In an age where the default approach is for people to yell at each other, Leith van Onselen prefers to bellow, as he did last week on the Macrobusiness blog with me as the target (https://www.macrobusiness.com.au/2017/11/henry-ergas-jukes-stats-falling-workers-share-2/). I don’t know van Onselen, but he describes himself as ‘unconventional.’ Never was a truer word said.
After all, a conventional economist, looking at this graph of the wages share:
would not think it was a continuous linear trend. Rather, there are clearly three phases, or if you want to use a more technical term, regimes: one that sees the share rise sharply beginning the late 1960s and then accelerating in the 1970s, propelled by Whitlam wages explosion and its after-signature; one, in which the wages share declines steeply, that begins as the wages explosion unwinds in the 1980s and continues up to the ‘recession we had to have’; and finally a relatively stable wages share since then, albeit one affected (as economists would expect) by the mining boom and by swings in terms of trade. That is why in my column I compared the current wages share to that which prevailed since the real wage overhang had been more or less corrected.
But that would be boring and conventional; instead van Onselen projects a straight linear trend through the entire period, from the earliest signs of the acceleration of wages growth to the present, and concludes there is a secular decline. Ignored is the fact that even the most rudimentary econometrics would reject the hypothesis of a stable trend; and it would also reject the hypothesis that the current share represented a change in regime from that which has prevailed since the late 1980s.
As if that were not enough, van Onselen also takes a decidedly unconventional approach to measuring profits. As I noted in the column, profit rates are not at highs, as the ACTU claims. van Onselen claims to rebut my statement but as best I can tell, he looks at gross profits—somewhat strangely deflated—without taking account of the fact that the capital stock doubled in the mining boom. Call me boring and conventional, but I prefer to use the national accounts data to derive a measure of the gross rate of return—see below; enough said.
With all that, it is certainly true that the wages share was substantially higher, and the profits share lower, in the 1970s. But few economists would regard that as a good thing. On the contrary, as I explained in my column, the wages explosion of that period priced many workers out of jobs—a point even Gough Whitlam and Clyde Cameron recognized (albeit too late). I would have thought that rather than lamenting the fall in the wages share from those levels, van Onselen would do better to caution against repeating the errors of the Whitlam government. But maybe that is too conventional.
A final, more technical point. There seems to be a tendency in this debate to rely on the quarterly national accounts. However, as the ABS notes in the latest annual accounts, the quarterly figures on income shares are subject to large revisions, especially when the terms of trade are changing. It is therefore far wiser to wait for the annual data before rushing into print.