Statistics New Zealand released a swathe of annual productivity data yesterday.
These annual productivity data focus on the so-called “measured sector”, whereas most often (for data availability reasons) productivity comparisons are done for the whole economy. The measured sector currently covers 77.3 per cent of the economy. It excludes ownership of owner-occupied dwellings, public administration and safety, education and training , and health care and social assistance – all sub-sectors where market price information is difficult or non-existent. The measured sector data are good quality but (a) are only available with a considerable lag (data released yesterday are up to the year ended March 2014), and (b) are mostly only useful for looking at New Zealand’s own performance over time (and only limited amounts of time, since the data on this measure go back only to the mid 1990s). Other databases, typically using whole economy measures, are more useful for timely cross-country comparisons.
The chart below shows measured sector labour productivity and total factor productivity growth since the year-ended March 1998. These measures don’t just use a volume measure of labour inputs (eg hours worked) but adjust for the changing composition (improving quality of the workforce). Simple measures based on hours worked in effect understate the role of inputs and, thus, overstate productivity growth.
On this measure, labour productivity growth does not look too too bad. In particular, although growth since 2007/08 has been slower than it was previously, the slowdown is less marked than many other series show for other countries. But bear in mind that over the 16 years shown, total growth in labour productivity was only 20.3 per cent – just under 1.2 per cent per annum.
By contrast, the TFP picture is sobering. In the 11 years since 2003, total TFP growth has been around 1.5 per cent (little more than 0.1 per cent per annum). As I’ve suggested previously, perhaps there is something in the notion that the higher terms of trade (since 2004) have undermined TFP growth, changing production patterns to take advantage of the higher output prices but in ways that reduced measured productivity. Perhaps, but I doubt if the effect can have been quite that large. And the sectoral TFP data back up those doubts. Here is the chart for agricultural sector TFP (only available to March 2013). It is a noisy series (droughts do that), but it looks as though there has been some TFP growth in the sector since 2003, unlike the picture in the aggregate TFP series.
Finally, a quick comparison with the Conference Board estimates for New Zealand, which I used in my series on cross-country comparisons since 2007. Here is the chart.
The Conference Board uses a model to estimate TFP which ascribes more of New Zealand’s growth to the growth in capital services (than SNZ do). (Like SNZ they make a correction for changing labour quality). There is no point directly comparing the number from the SNZ measured sector TFP series with the Conference Board TFP series – different models produce different results. But what is perhaps useful is to note that in both models New Zealand’s TFP growth has tailed-off markedly since 2003. That should be pretty disconcerting.
And here is the international context for TFP growth, with a focus on the post-2007 period.