Treasury reckons the deadweight costs of income tax means projects have to provide benefits in excess of nominal cost of somewhere around 20%. So if the proposed projects pass cost-benefit already by more than a 1:1.2 ratio, then they’re separable from asset sales. I’d expect that it’s only where the deadweight costs of asset sales are less than those from raising revenue from tax or borrowing and where the proposed projects only just failed cost-benefit at the 1:1.2 ratio that you can make a strong link between the two. I’d be a bit surprised if that were the case.

There are two reasons not to think of the two decisions as a portfolio. First, investing in schools, health, or whatever can be funded out of borrowing just as easily as from asset sales. The issue of whether the government’s net financial position is better with a portfolio of higher debt and higher ownership of equity (SOEs) is quite sepearte from whether that position should have more net debt coupled with social investment.

Second, it is a huge mistake of falling into the trap of calling social expenditure “investment”. Yes, in principle, expenditures in social areas can produce a return in the future, but, useful expenditures will be on-going (e.g. better teachers), not a one-off and the uncertainty bounds aroudn the return are too great to warrant financing out of borrowing rather than taxation.

My worry with National’s plan is that, in a silly attempt to make the asset sales more acceptable, they will put the money into bricks and mortar photo ops. Anyone with children at school knows that the most (only?) important determinant of a successful year is the quality of the teacher, not how nice the school hall is, how many computers the school has and whether there is broadband.

And, heck, I generally favour asset sales. But total sales, not partial. 

You’d hope so.

The way I see it – without any analysis there is an x% chance a policy will be bad.  They put two policies together, and the chance that the new one is bad is greater than 2x% – as they are already introducing arbitrarily political fiddling to the process.

Of course, if x>50% this equation might not hold perfectly …

Agreed – which gives it more than twice the chance to be bad policy!!!  Score.