MSD now only reports duration of benefit as less or more than 1 year so I put in an OIA to get the breakdown they previously published. I've graphed some of the data supplied. The first compares 2008 to 2013:
It's not a great picture. But perhaps to be expected given the GFC. The ten years plus dependency is a worry. With our rapidly ageing population the numbers long-term dependent should be moving onto Super at a faster rate than prior yet the number has actually increased slightly.
The height of the GFC effect was in 2010 so it's worth looking at what improvement there has been since:
As you'd expect the shorter term dependants decreased. But the 4-10 years number increased and the ten plus number has barely budged.
If I use the National government's official definition of long-term dependent - one year plus - the percentage has actually increased ever so slightly between 2010 and 2013 from 69 percent to 69.4 percent.