Thursday 11 September 2008

Inflation targeting and the medium term

This is a guest blog from Eric Crampton. Welcome Eric.

Eric is a Senior Lecturer in the Department of Economics at the University of Canterbury.

Alan Bollard today reduced the OCR by half a percentage point. While inflation remains well above the target band, he forecasts that it will dip below the top of the band sometime in the medium term (March 2010) and so he's fulfilling his obligations under the Policy Targets Agreement.

Just what constitutes the medium term? It's nowhere specified in the current Policy Targets Agreement. Most macroeconomists would deem the medium term to be the period over which monetary policy can be expected to have some effect, so somewhere from 18 months to 3 years, but RBNZ has been very coy about what it deems to be the medium term for purposes of the Policy Targets Agreement.

I'm not a macroeconomist, I'm a micro guy. And in microeconomics, we talk about revealed preference. Consequently, let's take a functional look at the medium term and define it as being the period over which we must average realized and expected inflation outcomes in order for the inflation rate to be within the target band. We're currently well outside of the target band and have been for a few quarters. Over how many quarters do we need to average in order for outcomes to be within target? If we only look backward, starting from September quarter 2008, we need to look back as far as December quarter 2003 in order for average inflation to be below the top of the target band. So, just short of 5 years. On a functionally-defined, backward-looking standard, the RBNZ deems the medium term to be almost 5 years. That's well beyond what most macroeconomists would deem a sensible medium term.

Now let's bring in forward-looking inflation projections. What happens if we center things on the September 2008 quarter and work out the number of quarters over which we must average, looking both forwards and backwards, in order for inflation outcomes to be within the band mandated by the Policy Targets Agreement signed by Dr. Bollard? Well, we quickly run into the problem that the projection ends at March 2011. If we look forward 10 quarters after the current quarter, and backwards ten quarters prior to the current quarter, the 21-quarter window has average inflation above the top of the target band. We then have a couple ways of proceeding. We can either then decide only to look backwards, or we can assume that inflation outcomes will remain constant after March 2011. In the former case, we then need to look back to September 2003 in order for inflation outcomes, on average, to fall below 3%. So, 31 quarters, or just short of 8 years, is the medium term. In the latter case, we hold inflation constant at the March 2011 rate (3%) for every subsequent quarter and find a medium term of 41 quarters: just over 10 years. If we're generous and assume inflation post March 2011 to be constant at 2.8%, the medium term drops to 39 quarters, or just shy of 10 years. We should worry too that RBNZ cautions that there's more upside risk to inflation going forward than downside risk: in other words, they're more likely to have underestimated future inflation than overestimated it. In that case, the medium term would be longer.

In the most generous case, the medium term is 5 years. In the more realistic case that includes inflation expectations, the medium term runs to about 10 years. Neither of these is consistent with my understanding of what constitutes the medium term as most macroeconomists would define the term, but I'm just a microeconomist.

We could, perhaps, chalk this up to there being something odd about the medium term in the antipodes. Time moves more slowly here, so we could hardly expect to finish a medium term in only a couple of years. On the other hand, if we apply the same method to defining the medium term to the period when Don Brash was subject to a 1-3% inflation target, the medium term never exceeded 7 quarters -- almost two years, or about what we'd expect under normal definitions of the medium term.

It could reasonably be argued that Alan Bollard has faced more difficult international conditions than did Donald Brash. And it's true: inflation in the traded sector has been much higher in the more recent period. But there's a big difference between the two periods. When inflation in the traded sector was high under Brash's tenure, inflation in the domestic sector was negative and consequently overall inflation rates only exceeded 3% for 3 quarters. It would have been malign to expect Brash to have done more during that period. By contrast, non-traded inflation has not dropped below 3% since Alan Bollard took office regardless of strong deflation in the traded sector in many quarters. Inflation outcomes under Bollard have only stayed under 3% in those quarters in which we've experienced low inflation or deflation in the traded sector. In short, it's only by grace of a benign international inflation environment that we've experienced any quarter in which Alan Bollard's RBNZ has been successful in meeting its target. And that environment has now changed.

Can a 0.5% cut to the OCR be consistent both with inflation outcomes expected to be beyond the top of the target band for another 5 quarters and with the RBNZ really trying to keep inflation outcomes within the target band over any reasonable definition of the medium term? Or, is it instead consistent with our needing desperately to look for a new Reserve Bank Governor after a change in government if we want to have any hope of getting inflation expectations back under control?

I find this all very depressing. In the country that's done more than any other to work to insulate its central bank from political pressures and to ensure bank independence, we've proven that a sufficiently determined government can completely erode that independence by appropriate eroding of the bright-line rules that characterized previous policy targets agreements, appointment of a dovish Governor, and sufficient winks and nudges about that there will be no consequence whatsoever to the Governor's ignoring of the target band. Indeed, what signals have come explicitly from Parliament are that, if Governor Bollard were to take the Policy Targets Agreement seriously, the PTA would immediately be replaced by a new one mandating that he focus on multiple targets rather than just inflation. With a wink and a nudge, an important part of our fiscal constitution is undermined to the point of irrelevance. And not to howls of outrage but rather to the thunderous applause of those who can't see beyond the next mortgage payment.

2 comments:

Anonymous said...

Yes, this should be very galling for any institutional economist. Here is the Reserve Bank's primary function of Bank from section 8 of the Reserve Bank Act 1989:

"The primary function of the Bank is to formulate and implement monetary policy directed to the economic objective of achieving and maintaining stability in the general level of prices."

If there is any dispute as to what trend in the measured rate of inflation is consistent with "stablility in the general level of prices" a sensible thing to do would be to ask the Government Statistician for an expert opinion on the question.

But does anyone really want to argue that the rates of inflation we are observing in NZ are consistent with "stablility in the general level of prices"? (I doubt that even the Governor is prepared to argue that.)

But if not, just what accountability is there for not delivering on the Bank's primary function?

Anonymous said...

The Board of the RBNZ is charged with figuring out whether the Governor is acting in a manner consistent with his obligations. Try emailing them and asking them what they think of current performance. I'll bet you'll find they're very happy with his performance: that they weigh the risks of the end of the financial world as being higher than a bit of piddling inflation, and that somehow a half or full point drop in interest rates can stave off an apocalypse. The "rules" provide ample scope for this kind of discretion given the lack of definition of the medium term, they would argue. But ask them yourself. Don't expect to be pleased with their answer.