Malcolm Edey’s General Taboo
No, the RBA is not financing the government
A story by John Kehoe makes much of Malcolm Edey’s ‘respectful disagreement’ with RBA Governor Phil Lowe as to whether the RBA’s purchases of government bonds for monetary policy purposes are financing government. In a recent addendum to an earlier paper on ‘The role of the RBA in Australia’s [global financial] crisis response,’ Edey suggests this goes against the ‘general taboo’ against monetary financing of government.
The disagreement is more semantic than real. Edey is not criticising QE, but rather seeking to clarify some of the confusion around the relationship between monetary and fiscal policy. Both Lowe and Edey acknowledge there is an indirect effect on the government’s cost of borrowing due to the RBA’s outright purchases of government bonds. In so far as QE lowers bond yields, it reduces the cost of government debt servicing.
As I noted in my USSC report on QE, however, that is only the static effect. The dynamic effect of QE should be to raise longer term interest rates. This was the consistent pattern of QE episodes in the US. Bond yield rose during rounds of QE and fell during the intervening periods because of changing expectations for growth and inflation.
Of course, a reflationary monetary policy will also grow the tax base, reducing the debt burden as a share of GDP. This effect is far more important. The effect of QE on debt servicing costs is a trivial issue in comparison.
This is no different from ‘conventional’ monetary policy implemented via the cash rate. Changes in the cash rate will affect longer term bond yields. Indeed, those yields are simply the expectation for the future cash rate, plus term and other premia. The government was publishing a yield curve assumption in the budget papers long before QE came along. So monetary policy will always have implications for debt servicing costs. QE may operate through somewhat different channels than the cash rate, but the substantive effects are the same and in no sense unconventional, much less ‘taboo.’ The effect is also cyclical. No one would suggest that the RBA is ‘defunding’ the government by tightening monetary policy.
It should also be recalled that the RBA’s balance sheet is effectively part of the Commonwealth’s balance sheet. The RBA pays a dividend to the government after deducting its costs, but that dividend is a by-product of its overall operations. As many a Treasurer will attest, the magnitude, not to mention the timing, of the RBA dividend can be a welcome addition to the revenue side of the budget. A dividend is not a given though. The BoE is not paying a dividend to the UK government this year for the first time in living memory and has even breached its capital buffers (try doing that with a private financial institution and see what happens to management).
Then there is seigniorage, the profit the government earns from supplying monetary liabilities in the form of notes and coins to the public at a face value in excess of their cost of production. An expansion in notes and coins on issue will increase seigniorage.
All of these things could be said to indirectly finance the government, but not in ways that are relevant to decisions about how we should use macroeconomic policy instruments, as John Kehoe’s story suggests.
Government bonds and base money are both government liabilities. If that sounds like MMT, well, it is. John Cochrane makes much of this equivalence in his fiscal theory of the price level, but government bonds and base money are also imperfect substitutes along a number of dimensions and not just liquidity. It is the failure to recognise this imperfect substitutability that confuses John, not to mention those well to his left. But I’m deferring judgement on John’s version of the FTPL until we get to see the final version in his forthcoming book.
QE could be implemented in a way that directly finances government spending, what George Selgin calls fiscal QE (and Edey calls ‘helicopter money’). This breaks down the formal separation of monetary and fiscal policy and is rightly criticised by Governor Lowe. However, David Beckworth has proposed a form of QE that is instituted by the Fed for monetary policy purposes, but relies on Fed-financed direct fiscal transfers rather than asset purchases as its operating instrument. This blurs the distinction between monetary and fiscal instruments, but otherwise maintains the institutional separation between the two arms of policy. Peter Tulip advocates something similar in a New Keynesian setting.
One confusion that arises on the part of some financial market participants is to look at government bond issuance as the net of issuance and RBA bond purchases. Government bond issuance is still an addition to the stock of debt, regardless of who purchases it in the secondary market. The new issuance may no longer be tradeable if bought by the RBA in the secondary market, but the debt is still out there. Even if held by the RBA to maturity, the government still has to eventually redeem it or roll it over with new issuance.
The institutions for the coordination of fiscal and monetary policy are obviously important, but it is not necessary for fiscal and monetary policy to work together for macroeconomic stability. There was no harm, for example, in the Howard government implementing tax cuts for supply-side reasons in the context of the mining boom at the same time the RBA was tightening monetary policy. If nothing else, the labour supply effects from lower taxes would help rather than hinder monetary policy, even allowing for a demand boost from the tax cuts. Similarly, there is no reason why fiscal consolidation should necessarily derail an expansionary monetary policy. Monetary policy can offset fiscal policy if it needs to, although the scope of that offset is probably smaller than most people imagine.
All of which is to say that monetary and fiscal interactions are complex, which is why we need to pay close attention to the design of monetary and fiscal institutions to ensure that they work in the way we intend. Making a ‘taboo’ out of monetary financing of fiscal policy is no more helpful than labelling QE ‘unconventional’ policy. QE is so conventional it is pretty much the textbook description of how monetary policy works, bearing in mind that most textbooks abstract from the institutional reality of monetary policy operating procedures for pedagogical reasons.
Finally, we should mention that central banks have their origin as financing mechanisms for government, most notably in time of war (see Vera Smith’s The Rationale of Central Banking). Politicians of yore recognised that monetary policy was a powerful instrument. Central banks were given independence because we didn’t trust politicians with those powers. If central bankers don’t do their job, politicians will take those powers back. When Governor Lowe tells us that monetary financing of fiscal policy will lead to inflation, politicians might just conclude that they should do his job for him.