We often hear that labour productivity is not growing fast enough and some argue it should be a focus of government economic policy. Yesterday, for example, Bill English was trying to blame the recession on the average growth in productivity under Labour.
However, economists like Brian Easton point out that labour productivity growth is heavily affected by the different quality of workers. Less productive workers can get jobs when employment is booming and are likely to lose them when the economy hits trouble. It is argued that these people entering and leaving employment overwhelm real changes in worker productivity in the labour productivity statistics. I wanted to find out whether labour productivity growth really gives a useful picture of the state of the economy, or whether changes in employment growth are dominant, masking any real changes in workers’ productivity.
So, I went and crunched the numbers (actually, I got this online calculator to do the hard part). I got the employment figures going back to 1986 and worked out the percentage of working age people employed each year, and then the rate of change between years. I got the labour productivity index figure for the same timeframes and calculated the rate of change each year for that too. Then I compared the rates of change in employment and labour productivity to see if there was any mathematically significant relationship. And there was. In technical terms, there is a negative correlation of -0.2. (0 means no relationship, 0.5/-0.5 is a very strong link, 1/-1 is a perfect link, more info here). That means when employment goes up labour productivity growth tends to be slower and when employment goes down labour productivity growth tends to be higher.
You can see the trend when you look at the graph of change in employment and change in labour productivity.
In the late 80s and early 90s, employment is falling and labour productivity growth is strong. In the last decade, employment growth was strong and labour productivity growth was weak. When employment plateaued and then fell in 2007-08, when labour productivity growth accelerated to its fastest in 9 years.
In many ways the labour productivity number is just a mirror of what is happening with employment. Greater productivity per hour worked is undoubtedly a good thing, but faster labour productivity growth across the economy is usually the result of fewer people being employed leaving only the ‘cream of the crop’. Indeed, several times in the last two decades total economic output and employment has been falling while labour productivity rose rapidly. Simply looking at labour productivity growth might lead to the conclusion that everything is going well, but a look at the wider picture shows it is not. In fact, in the 11 years from 1987 to 2008 in which the employment rate grew, labour productivity grew an average of 2.1%, while in the 11 years in which the employment rate fell labour productivity grew an average of 2.4% – 15% faster. Labour productivity only fell one year, 2007, which was the year of the second fastest rate of employment growth at the peak of a long employment boom.
This doesn’t mean we shouldn’t try to boost labour productivity, of course we should as long as it’s not done by cutting jobs. But it does mean that labour productivity is not a useful statistic to look at if you are trying to assess the health of the economy. Productivity is only one factor in production; focusing on it gives a misleading, often backwards, picture. Which is, of course, why National has concentrated on it. They could hardly attack Labour on better indicators of economic health like GDP growth, wage growth, or unemployment.
Just wait, I bet the labour productivity stats for 2009 will show fantastic growth, even though the economy is in deep trouble.