In making his case, Sumner argues that:
(1) Inflation targeting is flawed because it depends on inflation indices like the Consumer Prices Index. However, such indices are easily swayed by factors such as exchange rates and sales taxes – which may not require a monetary response. Moreover, inflation indices frequently misprice (or overlook altogether) housing. These issues, inter alia, make inflation indices unreliable as a primary target for monetary policy.
(2) NGDP targeting works because it addresses the dual concerns of macroeconomic policy – inflation and growth – with a single policy target. This prevents the policy incoherence and confusion associated with dual mandates. While it is technically true that the Bank of England’s Monetary Policy Committee has a single mandate (inflation) it is clear to most observers that they are also conducting monetary policy with a strong view to supporting the wider economy. An explicit NGDP target would be more transparent and increase the accountability of the central bank.
(3) Inflation targeting requires central banks to focus on consumer prices. This makes inflation targeting regimes prone to allowing asset bubbles to form when the economy is booming. These eventually destabilize the economy. Sumner does not suggest that NGDP targeting would completely eliminate this problem, but he does argue that it would impose more monetary restraint while the economy was growing strongly, thus discouraging widespread, credit-induced malinvestment.
The flipside of this is that NGDP targeting automatically leads to looser, or more expansionary, monetary policy during economic slumps than inflation targeting would allow. This, Sumner argues, prevents real shocks in one sector of the economy depressing other, otherwise stable sectors. In other words, NGDP targeting is a more effective economic stabilizer than inflation targeting.
Sumner illustrates this by pointing to the fact that countries such as Sweden, which employed an aggressive policy of monetary stimulus – as would have been indicated by NGDP targeting – suffered less severe recessions and stronger returns to growth than their peers.
(4) The Great Recession of 2007-09 was a smaller version of the Great Contraction of 1929-33. While many observers have assumed a simple financial crisis –> recession transmission mechanism, the actual pattern was far more complex, with lots of reverse causality. NGDP fell sharply before the worst of the financial crisis in late 2008. And because NGDP is effectively the total funds people and businesses have available to repay nominal debts, this led to widespread financial distress. Plummeting NGDP expectations then caused asset prices to fall sharply, adversely affecting the balance sheets of the major banks.
In other words, a localized crisis (sub-prime) was followed by a fall in NGDP, which led to financial distress, which in turn sent NGDP expectations even lower. That crashed asset prices, dramatically worsening the financial crisis, which then led to a deep recession. It was a vicious circle.
On the basis of this interpretation, Sumner argues that NGDP targeting would have (a) prevented such a large bubble building up before the sub-prime crisis, (b) prevented the bursting of the bubble from triggering a sharp fall in NGDP, and (c) prevented the financial crisis from worsening so dramatically in late 2008. To put it simply, while NGDP targeting would not have completely prevented the financial crisis or the recession, it would have made them far less severe.
So how exactly does Sumner propose that monetary policy be conducted in future?
• Firstly, he suggests that an NGDP growth target be set for the Bank of England. He does not propose a specific target, but indicates something in the region of 5 percent annual growth.
• Secondly, he suggests the Bank employ level targeting, which means targeting a fixed growth rate trajectory, and making up for any near-term shortfalls or overshoots.
• Thirdly, he suggests that the Bank target the forecast, including market expectations. Monetary policy should be forward-looking, not backward-looking as is the currently the case.
• Fourthly, following on from the previous point, he suggests that the Bank of England set up an NGDP futures market and subsidize trading of NGDP futures contracts. This would give monetary policy a compass: if NGDP futures prices rose, the Bank could tighten policy; if prices fell, they could loosen policy.
• In the long run, he argues that this market could eliminate the need for policy discretion. The Bank should promise to buy and sell unlimited amounts of NGDP futures at the target price, thus making their policy goal equal to the equilibrium market price.
• Essentially, this would mean the NGDP futures market forecasting the setting of the monetary base that was most consistent with on-target NGDP growth. The monetary base would respond endogenously to changes in money demand, keeping NGDP growth expectations on target.
Scott Sumner’s proposals do not necessarily represent the corporate view of the Adam Smith Institute. Nevertheless, the Institute believes very strongly that a serious debate about our monetary arrangements is long overdue, and is delighted to publish this report as a contribution to the public debate.