Is zero government debt desirable?

Tony Aspromourgos

This seminar addresses the fact that - due to a sequence of budget surpluses, but more importantly, actual and prospective privatisations - Commonwealth public debt could be extinguished in Australia in the next five years or so, and the current Treasurer would like to do that. It seems to me that there are five key issues raised by that possibility:

1. Is zero (federal government) public debt in any sense optimal? More particularly:

Is zero public debt desirable in principle?

Is zero public debt desirable in the particular Australian circumstances?

2. If the answer to the two questions under 1 (or at minimum, the second of these) is 'no', and Telstra is sold, are there alternative attractive Commonwealth asset or portfolio choices, rather than debt retirement? For example:

Reducing unfunded Commonwealth superannuation liabilities, and/or repairing the natural environment?

Or, in light of 1, running (appropriately sized) budget deficits.

Or even, not selling Telstra?

3. In the absence of a market for Commonwealth government securities (CGS), will the effectiveness of monetary policy implementation be compromised?

4. If the Australian government debt market did disappear, and government subsequently found it again had to begin issuing debt - which all previous modern history suggests is highly likely - will re-establishment of a market involve undesirable new start-up costs and difficulties?

5. Are CGS (held to maturity) a riskless or 'capital safe' asset (and over a large range of maturities), fulfilling a key role in many portfolios, and for which there are no adequate substitutes?

I may leave aside the last two questions, which will be addressed by another speaker - though with respect to number 5, it seems clear that in their own minds the authors of the government's discussion paper are concerned about a vacuum occurring in the role of the CGS market in asset pricing and, via the derivative CGS futures market, in interest rate risk management (Commonwealth of Australia, 2002, pp. 35-49).

Zero debt is not generally optimal

The discussion paper released by the Treasurer in October is a useful summary statement of the possible consequences flowing from extinguishment of government debt and closure of the associated market. But the deeper and prior issue is whether this should occur.

Consider the financial mathematics of 'sustainable' debt and deficits:

d = (g-i)D*

This simplified fundamental equation gives configurations of long-run or 'steady state' debt and deficit ratios - where d is the trend primary budget deficit as a percentage of GDP, D* is the chosen trend in the ratio of public debt to GDP, gis the trend GDP growth rate of the economy, and i is the average interest rate on public debt.1 To give it operational or intuitive significance, the equation can be read like this: given the trend growth rate of the economy (g), and given the trend in the real interest rate paid on government debt (i), government can choose a desired trend in the ratio of debt to GDP (D*). This then determines the required trend in the primary budget balance, d.

But how is the choice of D* made? Certainly if zero is the chosen value of D* then clearly d must be zero - and of course the overall deficit will be zero as well. The overall deficit, as a percentage of GDP, is just:

d+iD* = (g-i)D*+iD* = gD*

Only under the fluke that g = i will the desirable steady state value of d be zero, otherwise than in the case of D* = 0. And note that with D* positive, and assuming trend positive g, the overall public budget balance (gD*) is necessarily positive. In a sense this is just an expression of the proposition that the private sector needs CGS: in a growing system there is then a growing demand for government securities, and the only way the supply can be augmented is by the government running an overall deficit.

But one must be careful in interpreting this logic in one respect in particular. This equation is about outcomes or configurations along the steady state path - or long-run trends, if one prefers. Hence, for example, if a government starts with a value of D it thinks too high, and wants to move to a lower value of D, then in the transition that government would have to run budget deficits smaller than steady state size, and possibly budget surpluses (and/or possibly sell assets). In other words, this fundamental equation describes long-run financial ratios not transition paths between different values of D*. The feasible values of D* also will be conditioned by private sector portfolio preferences: what D* are acceptable to the private sector, and on what terms. At the end of the day, in a market economy the public debt has to be willingly held by the private sector. But the key moral of this brief commentary really is this: 'sustainable' public budget balances are consistent with a spectrum of D* values, of which zero is merely one special case.

Is zero debt optimal in the particular case of Australia? The essential issue in considering whether Australia in particular should have zero public debt is the trade-off between the benefits of increasing the trend value D* (and therefore increasing the trend overall budget deficit, gD*), and the costs of that. One cost might be a tighter primary budget balance, depending on the relationship between g and i. It may be noted also that higher D* might feed back upon and raise i. The latter is determined by the economically relevant characteristics of the asset (CGS). These asset characteristics (or at least the market perception of them - e.g., heightened perception of exchange rate risk) might well change as D* gets higher - though there are not likely to be any smooth continuous functions linking D and i, as employed in the appendix to the government's discussion paper.

Many issues could feed into a consideration of the costs and benefits of increasing the debt/GDP ratio. For example, the appropriateness of using more or less long-lived debt to finance capital expenditures is well understood and appreciated. (What more long-lived assets to invest in via debt are there than the repair of the Australian natural environment, to achieve a measure of environmental sustainability?) On the other hand - given the historic character of the Australian economy as a capital importer, and the associated high exposure to foreign liabilities - low public debt could be justified as a counterbalance to high private accumulation of foreign liabilities. In short, it is a complex matter, to which no glib answers need be given here. But it can be said, for sure, that it is extremely unlikely that the correct answer to what is the optimal value of D will turn out to be the current Treasurer's very 'neat' solution, zero. With Australian Commonwealth debt now so low relative to GDP, it's hard to imagine real benefits from going lower - other than political symbolism for the electoral advantage of the incumbent conservative government.

What about alternative asset accumulation, rather than debt elimination, if Telstra is sold?

In fact, from the point of view of 'rational portfolio choice', in the choice of portfolio of assets and liabilities by the Commonwealth, Telstra should only be sold if there are more desirable assets for the Commonwealth to acquire, or liabilities for it to reduce. In a nutshell: should the government's portfolio be shifted to 'no Telstra' - and if so, should the new portfolio involve no debt, or should it rather involve other assets (e.g., assets to better balance superannuation liabilities of the Commonwealth, or enhancement of natural assets like Australian waterways)? Again, no pat answer need be provided here; but properly, it is with a view to the trade-offs between these possibilities that the matter should be deliberated upon. Certainly, to preserve debt and use Telstra sale proceeds to fund environmental public works would amount to saying, in a sense, that long-lived natural asset enhancement warrants long-lived public debt.

What are the implications for monetary policy?

The discussion paper is remarkably brief on the implications for monetary policy (pp. 54-6, 62). This is a technical question to which I would not proffer an emphatic answer. But one would expect that as a matter of principle the Reserve Bank of Australia should make a submission to the Commonwealth Debt Management Review, due by 6 December (p. 13), and publicly release its submission - as a public statement of its view concerning the implications for monetary policy (and perhaps also concerning the option of the Commonwealth acquiring an asset portfolio, since international experience, as outlined in the discussion paper appendix, makes clear that the RBA would likely have a key role in the management or at least supervision of any such portfolio).

Conclusion

Might there be some politics involved? In conclusion, one may wonder whether this whole proposal of closing the debt market might not be just a kind of large-scale government publicity stunt, though others who are closer to the politics will be better able to judge. The debate around the issue has of course gained considerable (free) publicity for the government's debt reduction outcomes - outcomes which seem to be a positive for the government with 'average punters'. Or, in a somewhat similar vein, perhaps there is a measure of ambit claim in the proposal, with a Telstra sale eventually being used for a mix of debt reduction, environmental and other public asset rebuilding, and perhaps asset accumulation in support of Commonwealth unfunded superannuation liabilities. P. Costello, Treasurer, might be of sufficiently limited imagination to be attracted to leaving a 'monument' like zero debt; but P. Costello, PM, might be more alive to the electoral attractions of, for example, funding an environmental policy that would draw votes from RARA (rural and regional Australia) and preferences from Green voters - and so might J. Howard, PM, be similarly alive to those attractions, if he stays on.


Tony Aspromourgos is Associate Professor in the School of Economics and Political Science at the University of Sydney. This a revised version of the text of an address to the Evatt Foundation Breakfast Seminar - Who's afraid of public debt? - at the Southern Cross Hotel, Sydney, 19 November 2002.


Notes

1. It is difficult briefly to provide an intuition of this equation, but consider this: starting at the desired debt ratio D*, to keep that ratio constant debt must grow at the same rate as GDP. So, write an equation equalizing the growth of debt and GDP. Manipulating such an equation then produces the equation above. It should not surprise that this equation includes d, i and g: the first two determine the growth of debt; the third determines (in fact is) the growth of GDP. (The 'primary' budget deficit, by the way, refers to the budget deficit net of interest payments on government debt. The budget deficit as a whole, as a percentage of GDP, is d+iD*.)

Reference

Commonwealth of Australia (2002), Review of the Commonwealth Government Securities Market: Discussion Paper, Canberra, ACT (ISBN 0 642 741662).

See also:

Suggested citation
Aspromourgos, Tony, 'Is zero government debt desirable?', Evatt Journal, Vol. 2, No. 8, December 2002.<http://evatt.org.au/papers/zero-government-debt-desirable.html>