Go back to John Quiggin's home page









Centre for Economic Policy Research,

Australian National University






Quiggin, J. (1995), 'Does privatisation pay -- A reply to Domberger', Australian Economic review 95(2), 48-49.




I thank Simon Domberger for his comments, which go a long way towards identifying areas of common ground and isolating our key area of disagreement. Before turning to this key area. I would like to deal briefly with some peripheral issues.

On underpricing, Domberger makes the important point that, like public ownership, privatisation may be used to pursue political objectives at the expense of net economic benefit, and that this partly explains the grossly inflated discounts of the Thatcher period. I agree with his conclusion that the grossly inflated discounts of the Thatcher period are a thing of the past, but would caution that some degree of politically motivated discounting is likely to be a continuing feature of public floats.

On the issue of fiscal illusion, I meant to refer to reductions in government saving arising from impacts on the measured Budget deficit, rather than reductions in private saving as Domberger's comment suggests. My claim is that at least some of the proceeds of privatisation (as measured in the Budget) may be used by governments to finance a higher level of expenditure, or a lower level of taxation, than would otherwise be the case.

Thirdly, on deadweight losses, Domberger argues that my treatment is asymettric, since I do not take account of the benefits of reduced public debt interests. I have shown that the stream of profits foregone is larger, on average, than the stream of interest saved. A symmetric analysis implies that there is a deadweight cost associated with this net loss. This is most important in cases where the loss arises from underpricing and might be mistakenly regarded as a (costless) transfer.

The critical issues raised by Domberger relate to the equity premium and to the appropriate charge for risk. Domberger argues that the government bond rate may be misleadingly low because the associated risk is borne by taxpayers. But the equity premium puzzle is similarly difficult to resolve if attention is confined to high-quality private debt. The basic problem is that the premium demanded for bearing systematic risk is far higher than would be expected from a perfect capital market, on the basis of plausible assumptions about consumer preferences.

It is necessary to distinguish here between pure risk, in the sense of variability about a given mean and 'downside' risk associated with the possibility of adverse events. Downside risk is relevant, for example, when a project is evaluated on the assumption that construction will proceed to plan or when a bond is evaluated at face value without considering the possibility of default. Pure risk reflects the premium that would be paid for a fixed income stream as opposed to a risky income stream with the same expected value.

At least in principle, the methods proposed here take full account of downside risk, particularly in the evaluation of cases where privatisation was proposed but not undertaken. In this case, it is possible to observe the actual stream of profits that would have been foregone. In the cases where privatisation was undertaken, the assumption that real profits would have been maintained after privatisation appears conservative in view of the massive increases in profits that typically took place. More generally, the issue of whether the estimates of average reductions in public sector net worth associated with privatisation presented here are representative is an empirical one, which may be resolved by the collection of further evidence.

The specific example of banking is worth examining in more detail. In my paper, I considered the proposal put forward in 1985 to privatise the Commonwealth Bank and concluded that a sale price of $2.5 billion (1993/94 values) would have been offset by the loss of a stream of earnings with a present value of $10 billion (using a 5 per cent discount rate and taking the 1985-94 period as representative), implying a loss of $7.5 billion.

As Domberger points out, my analysis does not include cases of large public sector losses such as those of the State Banks of Victoria and South Australia. Suppose that, in addition to the Commonwealth Bank, the privatisation proposal for consideration was expanded to include the various State Banks. In view of the small size of these banks in 1985, it is unlikely that they would added much more than $500 million to the sale price. Furthermore the results derived from the Commonwealth Bank case would probably extend to profitable state enterprises such as the State Bank of NSW. However, the public sector would have avoided the huge losses of the State Bank of South Australia (now estimated at between $2 billion and $3 billion), and the smaller losses of the State Bank of Victoria. Large though these figures are, they are still smaller than the loss that would have resulted from privatisation of the Commonwealth Bank in 1985. Thus, complete privatisation of the public banking sector would have reduced public sector net worth, as estimated using a 5 per cent discount rate.

Furthermore, it is important to remember that privatisation would have only partly insulated the public sector from the losses incurred by the State banks. The guarantee of deposits offered by the Reserve Bank has incentive effects similar to those associated with full public ownership. The difference is that if a private bank is closed down, the losses fall first on the shareholders and only then on the public sector guarantor. However, the costs of this process to the guarantor are such that in practice, some form of rescue is preferred in all but the worst cases. It seems likely that a privately owned State Bank of Victoria would have been rescued by some form of merger similar to the one that actually took place. The State Bank of South Australia was beyond rescue. However, its losses were so large that they would certainly have exhausted the equity of the shareholders, leaving a large proportion of the loss to be borne by the public sector guarantor.

In summary, even taking into account disasters like those of the State Banks, the average flow of profits foregone in privatisation is greater than the average saving in public debt interest. The key remaining issue, resolution of which requires further theoretical analysis and more empirical evidence, is whether this reduction in the average value of public sector net income may be justified as a pure risk premium.


As Simon Domberger says, this issue must be assessed on a case-by-case basis. Such assessment has been sadly lacking in the past.

Go back to John Quiggin's home page