Tuesday, April 23, 2013

War on Multipliers

The war on multipliers has scored a battle victory in the NSW Land & Environment Court. Peter Martin has posted on a decision which overturned approval for a Rio Tinto coal mine in the Hunter Valley:  Really Rio? The judge who put its claims about jobs to the test

This is an excellent expose on the Loose Change theme of dodgy consultants multipliers with Peter Martin also posting the court transcript of Richard Denniss v Andrew Searles. In this case Searles is the consultant for Hunter Valley Research Foundation who identified a massive employment multiplier for the Rio Tinto project.

This isn't new and a Senate estimates question last year almost perversely from a Greens senator drew a response from the current Treasury Secretary that there is no industry employment multiplier when an economy (currency zone?) is close to full employment. Resources are rather transferred.

Denniss has estimated that input-output industry multiplier models would employ 200% of the labour force. I think that maybe conservative at a regional level where consultants reports for the Far North economy with multipliers for everything from the Bruce Highway to the Gordonvale CWA sewing bee should easily employ at least 200% of the labour force and eliminate unemployment entirely!


  1. This has been a long-standing problem with all projects, public and private. The Paul Freebody proposed waterpark is said to employ 28,000 during construction.

    Perhaps a standard metric like man-days should be mandated.

  2. It could be argued that the FNQ unemployment rate indicates available labour resources and that the region is too small anyway to impact policy in the currency zone. However that hasn't stopped the abuse of multipliers or considerations of structural unemployment in FNQ. Perhaps Warren Entsch should abandon the idea of Cairns as a tropical campus of AIS and instead champion a centre of expertise in regional economics?

  3. One of the OESR boffins wrote a great note on the limitations of IO analysis back in the early 2000s that is worth a read: