Friday, May 21, 2010

More on the RSPT

I read a couple more papers on resource taxation and am not much clearer about things. Ben Smith wrote about the impossibility of a neutral resource rent tax and Diderik Lund wrote
a recent review.

It is pretty clear that a pure "Brown tax" where the mining company immediately gets refunded the tax rate (say 40%) multiplied by all losses minimizes the effect on investment decisions as long as the government is definitely going to keep on doing that. The government becomes an effective passive shareholder in 40% of each project. There is still a question of the government getting a free ride on the intangible capital/human resources of the mining company, which aren't usually expensed to individual projects. Under certainty, this would then be a neutral tax on investment.*

Further complications ensue when losses must be carried forward or some expenses are excluded. The RSPT carries forward losses at the long-term bond rate because the Henry review argues that as the government will eventually refund their share of any terminal loss the company is effectively lending money to the government. Whether this results in a tax that is still neutral in its investment effects is debatable, it appears to depend on modeling assumptions. Investors might not mind lending money to the government at the government bond rate but it is a different question to force them to do so when they have other investment opportunities. Emphasizing this point, the tax doesn't allow the deduction of interest expenses. The company's existing cost of capital is likely to be higher than the government bond rate. The Henry review lists a number of other expenses that cannot be deducted some of which seem reasonable and some not. All these variations on the pure Brown tax certainly reduce the neutrality of the tax.

I just read comments in the Australian from Ross Garnaut which further explain the reasoning behind the use of the government bond rate. If the government is promising to repay all losses then banks lending to the mining company should also be prepared to lend at the government bond rate. So, by this thinking there is no wedge between the returns on the RSPT capital account and the company's cost of capital. Garnaut is saying that this is a simplification which isn't likely to hold in reality. It certainly doesn't apply to existing projects - but accelerated depreciation is meant to deal with that issue. Shareholders will effectively put in 60% of the capital to projects and get the normal rate of return on that 60%.

I think I understand fully where the Treasury is coming from now.

* As Jerry Hausman argued after I wrote this blogpost originally things get more complicated even for this best case Brown tax when there is uncertainty.

No comments:

Post a Comment