Measuring Productivity

I'm sure Robert Gottliebsen had a merry chuckle when writing in his spray on the public sector today that "Cynics say the current definition of productivity in Canberra is related to lifting staff numbers and expenses."

In fairness he went on and said "[Treasury secretary] Parkinson must move outside the public service barons and find simple ways to measure government productivity. It will not be easy."

He then flays around looking for 'measures". The first is a simple intent measure - how much does self-reporting show a gap between saying productivity is important and having a project to do something about it. he then holds up two miners as places where productivity is well measured.

Parkinson's Minister is responsible for both the Australian Bureau of Statistics and the Productivity Commission. I don't think he has a shortage of measurement capability at his disposal.

What he does have is a definitional one. Productivity measures how much output per unit of input. If you are producing a commodity measuring output is relatively easy because it is well defined.

If you are producing "policy" it is much harder to judge. Is a productive department one that sends out lots of discussion papers per employee. Is the ABS best measured on the number of excel spreadsheets issued per month?

The simple answer is no.

I've already criticised BOF for his argument about public sector wage freezes. He has used simple (labour) productivity measures of students per teacher, nurse per hospital bed. But productivity here will be measured in the quality of the output not its quantity.

Of course, productivity in its widest sense should mean total factor productivity, not labour productivity, but that itself is another issue for another day.

The worrying fact is a new agenda being waged from the right/business community that asserts that there is productivity improvement required in Government - and to make that an excuse for its own inefficiencies.


Novae Meridianae Demetae Dexter delenda est