## Sunday, November 11, 2012

### Time enough for success

Skepticlawyer's excellent post on the GFC examines the financial crisis. The post below is concerned with the time period for monetary policy. While, as I note below, the collapse in total spending clearly worsened the GFC, this post is more about how to avoid or minimise recessions and, particularly, events such as the Great Recession.

The remarkable success of the Australian economy in avoiding a recession (defined as two consecutive quarters of negative economic growth) for 21 years and counting has been remarkably ignored by economists and policy-makers across developed countries.

It is true that the robust growth of the Chinese economy has been good for Australian commodity exports (the usual reason for dismissing Australia's recent economic success). However, it is also true that the high $A has been less happy for other sectors of the Australian economy -- such as tourism, education and manufacturing exports. Moreover, commodity prices dropped dramatically during the height of the Great Recession -- Australia had a higher (proportional) drop in exports than did the US. The Australian success in avoiding recessions also predates the recent surge in its terms of trade. (Graphs in this post taken from here.) Like other market monetarists, I attribute the Australian success in avoiding those transaction crashes we call "recessions" to the Reserve Bank of Australia's (RBA) monetary policy. This is not to deny that decades of reform have made the Australian economy much more flexible, and so responsive to changes in economic conditions. Or that the concern of fiscal authorities to make life easier for the RBA has not been helpful (at the very least, in keeping down debt liabilities of taxpayers). Furthermore, as Jim Belshaw points out, the ability of the$A exchange rate to respond rapidly to an economic shock was very beneficial to the Oz economy during the GFC -- but that was also true during previous economic shocks, such as the 1997 Asian crisis.

All these things have made it easier for the RBA to run monetary policy smoothly. But it is that policy itself which is at the core of the avoiding of transaction crashes.

The business cycle itself has not been abolished -- this is particularly obvious if one looks at per capita GDP data.  It has, however, been greatly, and beneficially, ameliorated.

At the core of market monetarism is the view that monetary policy is not a mechanical manipulation of monetary aggregates or base interest rates but an exercise in managing expectations through policy signaling. (A useful, quick, lay-friendly summary is here; I would also add in debt as the ultimate "sticky price".)

Within the boundaries set by the level of credibility of the central bank -- credibility that may require commitment to action to maintain --  the same policy instruments may have quite different signaling, and so expectation management, effects depending on the policy framing in which they are embedded.

Balanced credibility
As I have argued elsewhere, the advantage the RBA has is that, unlike central banks who only have credibility on inflation targeting, the RBA also has credibility on total spending (and so income) in the economy. Its credibility is balanced (inflation and spending) not unbalanced (inflation only). Which means that its interest rate shifts positively manage both inflation and spending expectations.

Confidence that spending (and so income) will remain relatively stable means that Australia has not experienced the transaction crashes we call "recessions".

How does the RBA achieve this balanced credibility? It is explicitly an inflation targeting central bank and it uses base interest rate as its policy instrument; its policy signaling device. In this, it is like other central banks.

The key difference is the time period of said inflation target -- it is an average over the business cycle. In the words of the RBA website:
The Governor and the Treasurer have agreed that the appropriate target for monetary policy in Australia is to achieve an inflation rate of 2–3 per cent, on average, over the cycle. This is a rate of inflation sufficiently low that it does not materially distort economic decisions in the community. Seeking to achieve this rate, on average, provides discipline for monetary policy decision-making, and serves as an anchor for private-sector inflation expectations.
As I have also argued before, this means that its policy time-horizon is based on economic conditions, not on some time period imposed over the top of economic conditions. To put it another way, there is no time period constraint operating on the RBA (and its policy signaling) independent of economic conditions.

Since the policy time-horizon is directly connected to economic conditions, this greatly helps signaling (and so expectations management). There is no concern that some arbitrary (time) constraint unconnected to economic conditions will affect RBA policy. A time-horizon based on economic conditions does not impose an arbitrary -- and so potentially dysfunctional -- constraint on policy.

Moreover, because it specifically invokes the business cycle, the RBA's monetary target actively assists maintaining balanced credibility. If the inflation target is an average over the business cycle (as it has to be to be responsive to economic conditions), that clearly implies more strongly than a simple inflation target that monetary policy will be easier if economic activity weakens and tighter if economic activity strengthens; not merely to "keep" the inflation target but in order to "lean against" the direction of economic activity so as to stabilise spending. For, being an average, it clearly implies the inflation target constraint will be (temporarily) traded-off to keep economic activity (or, more accurately, since central banks only directly control matters nominal -- that is in money terms -- spending) up if economic conditions worsen, thereby creating both inflation and spending (and so income) credibility, i.e. balanced credibility, for the RBA. So its interest rates shifts provide credible and positive signals for both inflation and spending.

Supply shocks, money demand and looking forward
This also makes the RBA somewhat more broadly forward-looking than other central banks. Since the target is a band over the business cycle, how inflation has been is considered in terms of expected economic conditions, particularly if economic conditions are weakening. Scott Sumner nicely highlights an example of this.

To put the difference between simple inflation targeting and (implicit or explicit) spending targeting another way, as Scott Sumner points out in his recent (very clear) paper on NGDP targeting (pdf: nominal GDP = GDP in money terms = total spending on/income from output of goods and services), simple inflation targeting responds quite differently to supply shocks than NGDP targeting or some implicit spending targeting. Inflation targeting would lead to tightening monetary policy in response to a negative supply shock, such as a surge in the price of oil (a perverse response to expected conditions making a transaction crash much more likely), while NGDP targeting would lead to easing monetary policy (the correct response if a serious transaction crash is to be avoided).

Supply shocks show up how the RBA monetary policy time-horizon makes inflation-targeting operate like NGDP targeting, for the average-over-the-business-cycle-goal does not sacrifice the overall level of spending to the inflation target if economic conditions weaken due to a supply shock. A similar point operates when there is an increase in demand to hold money (driven, say, by a financial crisis). An inflation-targeting central bank is likely to be limited in its monetary response, as any effect on inflation from an increase in the demand to hold money is likely to be downward. A NGDP-targeting central bank would ease, since money being held is not being spent and so would have a serious downward effect on spending.

Given the US$is the premier global reserve currency, this makes it more important, not less, that the US Federal Reserve target NGDP (or else, like the RBA, has an explicit average-over-the-business-cycle target) since monetary-demand-shocks (for the premier global reserve currency) are more likely and the consequences of a US transaction crash are more serious. The "passive tightening" of the US Federal Reserve during the surge in demand for$US helped make the GFC such a financial crisis (as spending, therefore income, expectations weakened dramatically, worsening fears over debt and leading to a flight to cash) and, through the consequent transactions crash, the Great Recession "Great".
Regarding the time-horizon for policy, there is also an issue with what is known as level-targeting, whether of the price level or of NGDP.

Level targeting anchors longer-term expectations by forcing central bank to, if it follows the target, to adjust to past outcomes or be increasingly exposed as failing to meet its target. It does generate a possible short-run problem if reaction to past outcomes is held to create perverse responses to expected conditions. This is much more a problem for price-level targeting than NGDP-level targeting because the former lacks spending (and so income) credibility and the latter does not.

Repeatable success
So, the central bank having a policy time-horizon which is based on economic conditions allows much better policy signaling and expectations management. Having policy targets whose time periods impose arbitrary constraints unconnected to economic conditions on monetary policy is unfortunate. Having credibility on inflation but not spending makes serious transaction crashes not merely likely but as inevitable as anything can be in economic policy.

The success of a monetary policy target whose time-horizon is based on economic conditions and which has credibility for both inflation and spending is not some happy accident. It is a repeatable success open to any central bank which can bear to learn from antipodean achievement.

POSTSCRIPT: It is sometimes suggested that the RBA has been lucky in that it has not had to confront the problem of the "zero bound" (when base interest rates are 0%pa and so cannot be cut any further: the various quantitative easings [QEs] are attempts to get around that constraint). Australia has persistently run slightly higher inflation than, for example, the US.  Given the long history of Australian labour market rigidity (i.e. strongly "sticky" wages), this seems to be fairly clearly deliberate policy, fortuitously or deliberately making it much less likely Australia would confront the problem of the "zero bound".