Tuesday, 16 March 2010

Canadian Economics 3

I'm following up on Eric's call for macro commentary on Nick Rowe's liquidity trap comparison of Canada, New Zealand and Australia.

In brief, Nick notes that Canada started the recession with low short-term interest rates, and responded by reducing them as far as possible without hitting the zero-interest lower bound, whereas Australia and New Zealand started with much higher interest rates and were thus able to reduce rates a lot more. But while Australia fared better than Canada, New Zealand fared worse.

My central banking days are long behind me, and I try not to think about macro, but I'll give this one a go. I see two relevant facts.

The first thing to note is that manipulation of aggregate demand is a very limited government tool that can achieve two things: During stable times, small tweaks in interest rates can be a useful way of keeping the economy growing at potential without large swings in inflation or output. And after a major, real, hit to the economy, they can be a way of preventing contagion across the economy in which a downturn in one sector leads to reduced demand in another, and so on. In this case aggregate demand stimulation serves to stop a bad situation becoming worse but can't offset the intial shock.

My sense is that the recession came to New Zealand as a fall in the commodity prices that matter most here, and that while monetary policy could prevent contagion from that, it couldn't change the reality of a decline in our terms of trade. Canada, in contrast, which is much more integrated into the U.S. economy, sufferred contagion from the U.S. largely addressed by U.S. monetary and fiscal policy, with only a small contribution needed from the Bank of Canada. I don't know much about Australia, but it is important to note that the New Zealand and Australian economies are much more different than people think, in particular because Australia derives much more of its income from mineral wealth and so depends on a different set of commodity prices.

The second point to make is that the overnight interest rates that central banks control have only a very limited impact on aggregate demand. They are a very good tweaking tool for making continuous small adjustments to keep the economy stable, but even large movements in overnight rates won't have much impact on demand unless they are perceived to be long-lasting and hence are able to move medium-term rates. (To paraphase a comment about reserve asset ratios from my undergraduate lecturer, Frank Tay, trying to control aggregate demand with frequent changes in overnight interest rates is like an impotent man lifting his trousers up and down.)

By this view, a small reduction in interest rates in Canada close to what is seen as an absolute lower bound, would be viewed by the markets as likely to last for a long time (on the assumption that rates would have been lower if possible), and so even a small fall in short-term rates would be likely to lead to falls further out the yield curve. In New Zealalnd, however, starting from higher overnight rates, a similarly small fall without a credible reason for not reducing further would have generated expectations of a shorter-term policy.

It would be easy to check this theory by looking to see which country witnessed the greater falls in interest rates further along the yield curve, but I'm going back into macro retirement. I'll look forward to seeing if Nick follows up on this.


  1. Hi Seamus! It is great to meet up with you again, even if via the internet. And it's great to see you blogging.

    Your terms of trade story might indeed be part of the explanation of differences between Canada, Australia and New Zealand.

    The Bank of Canada did indeed make a "conditional commitment" to keep the overnight rate low until after Q2 2010, unless there were definite signs of inflation. That might have helped.

    Looking at longer term interest rates creates an ambiguity though. Low long term (nominal) interest rates might be interpreted as evidence of "loose" monetary policy, but they might also be interpreted as evidence of the low expected inflation and low expected real growth that are the symptoms of "tight" monetary policy. That's the trouble with the new-fangled Neo-Wicksellian approach of thinking of monetary policy solely in terms of interest rates.

    I wish there were more "comparative analysis", to try and help us see what worked and what didn't.

  2. Don't forget that NZ was in recession before the crash - previous govt polices had already tanked our economy. So the position that "NZ fared worse" might not be true, if we assume our "normal" state is the state we were in before the crisis. That "normal" state was recession....

  3. Nick: LIkewise, good to catch up again.

    Yes, it is a pain to disentangle the relative importance of expected inflation and expected monetary policy on medium-term interest rates, but I can think of a couple of possible ways.
    1. Subtract off expected inflation using an independent measure of it, derived from inflation-indexed bonds (if they exist in sufficiently thick markets), consensus forecasts, or something.
    2. Have a look at changes in the horizon inbalance in banks borrowing and lending to get a feel for how much of changes in medium-term rates are due to banks changing their willingness to borrow short and lend long.