Has the Fair Work Act made the labour market less efficient at matching unemployed people to jobs?  One way economists would try to answer that question is with the Beveridge curve.

The Beveridge Curve shows the relationship between the unemployment rate, along the horizontal axis, and the job vacancy rate, along the vertical axis. You generally observe a stable relationship between these two things – when unemployment is high, vacancies are low, and vice versa. As the state of the economy improves or worsens, we move along the curve.

If the Beveridge curve moves outwards, that’s usually taken as a sign that the labour market has become less efficient. An outward shift means that there is a larger number of unfilled vacancies per unemployed person  at every given level of unemployment. Confronted with an outward shift in the Beveridge curve, most economists would conclude that ‘structural unemployment’ has risen.

The following chart shows what has happened to the Beveridge curve since 2000. Each dot represents a month. You can see that from 2000 to mid-2009, the labour market remained on a stable Beveridge curve, but since 2009 the curve has shifted outwards.

Beveridge curve: 2000 to 2013

bev curveSo, is this evidence that the Fair Work Act has made the labour market less efficient? No, it isn’t. If you’d looked closely at the chart above, you’d realise that that it doesn’t show the Australian Beveridge curve at all – it depicts the US labour market. The US Beveridge curve just happens to have shifted outwards at around the time that Australia adopted new labour laws. I may be wrong here, but I don’t think even the staunchest critics of the Fair Work Act would blame the Australian legislation for developments in America.

I have posted this for two reasons. First, as a little reminder about correlation and causation. Everyone knows that correlation doesn’t equal causation, but that’s often forgotten in the heat of making some argument or another. If the graph above was indeed of the Australian labour market, there would have been a glut of op-eds pointing to this as evidence of the legislation’s pernicious influence on the economy.

The second reason for posting this is to register a note of discomfort about the way the Beveridge curve is sometimes used. I think it’s a very useful tool for visualising the relationship between two key labour market variables, but I think economists can be too quick to look to labour market institutional arrangements to explain shifts in the curve. Recessions make a lot of people unemployed and some of those people will remain unemployed for a long time. The longer they’re out of work, the less likely they are to be hired – this shows up as an outward shift in the Beveridge curve. But the key policy implication of this, in my view, is that we should be extremely concerned about avoiding recessions, as well as getting out of them as quickly as possible if we do find ourselves in a hole. A Beveridge curve shift need not have policy implications for labour market regulation itself.

In case you’re wondering, here’s the Australian Beveridge curve over the same period. The latest observation, for the February quarter this year, has us right around the trend line for the past couple of decades.

Australian Beveridge curve: 1992 to 2013

Aus bev curveNote: The ABS Job Vacancies survey was suspended between May 2008 and August 2009, so the quarters in between are not able to be represented on this chart.