Nolan on asset sales

Matt Nolan has an interesting piece on Asset Sales up at the TVHE blog. In his article Matt writes,

In New Zealand at the moment there is definite scope for opening up SOE’s to private sector investment – that is where we are sitting now. However, even given this I cannot go as far as Roger Douglas and say that the price does not matter – in fact, price is THE issue that the government should use when deciding whether to sell assets.

and adds

At the same time the government know that, if it keeps hold of the asset, it expects to make some dividend yield from said asset through time. As a result, the government can price the asset – they can say they would not accept a bid below the discounted expected return from holding the asset.

What I would argue is that the government may  - in some circumstances - want to accept a bid below the discounted expected return from holding the asset. One example as why it could want to do so is given by Anbarci and Karaaslan's idea of An Efficient Privatization Mechanism:

In this paper, we consider the privatization of State-Owned Enterprises (SOEs) that are legal monopolies but not natural monopolies; their markets can be opened to competition once privatization takes place and other competitors can emerge and compete successfully against them in a few years. But until that happens, these privatized SOEs can have a significant level of market power. The currently used “Revenue Maximization (RM)” privatization scheme maximizes the government revenue from privatization but does not provide sufficient incentives for the privatized SOE eiher to charge a price lower than the monopoly price or to improve production efficiency until competition arises. We propose a new scheme to privatize such SOEs. We term this new scheme the “Welfare Maximization (WM)” scheme. The WM scheme practically yields no revenue to the government from the privatization of any such SOE; however, it induces the privatized SOE to charge a competitive price in the absence of any regulation. It also turns out that the WM scheme provides greater incentives for post-privatization process invention (i.e., for post-privatization cost reduction) than RM scheme. (emphasis added)

This is a very specific situation but it helps make the point that just trying to maximise the price received for an asset is not necessarily a good idea. In the above example welfare is maximised while revenue is basically zero. So I would say that the price doesn't matter or at least the price is one of the least important factors in privatisation. The issue when thinking about whether to privatise or not is not price, but productivity. It is more important to get the regulatory environment right so that competition can breakout in the industry than it is to maximise the price for which the asset is sold.

Basically, I guess, I'm arguing we should have lexicographic preferences, with price low on the list.