Why don't we have disinflation (until now)?

Just a quickie, between meetings.

Paul Krugman asks "Why don't we have deflation?" His answer is: downward nominal wage rigidity. And he shows that graph of frequency distribution of nominal wage changes with a big spike at zero.

I don't think that's quite the right question. So I don't think that's quite the right answer. (Even though, sure, that graph is almost certainly telling us something important.)

"Deflation" means falling prices (a negative inflation rate). "Disinflation" means a falling rate of inflation, so prices rise less quickly than they did before. My question is: why don't we have disinflation (until now)?

I used to think I understood the Phillips Curve at a shallow descriptive level, even though I knew I didn't understand it at a deeper explanatory level. Now I think I don't even understand it at a shallow level.

I'm going to talk about Canada, because: I'm Canadian; and because Canada is simpler, because everyone knows what the Bank of Canada is doing and because Canada had a recession but didn't really have a financial crisis. But I think that what I say about Canada still applies, more or less, to other countries.

Look at the graphs of CPI and core CPI in this post, updated in this post.

During the recession, total CPI jiggled around a bit, so it was sometimes above and sometimes below the 2% target, but core CPI kept on rolling. There was no disinflation. Until very recently, when there finally does seem to be disinflation. (And January's CPI data that came in after those posts confirm my suspicions.)

If you had magically shown me those graphs in 2007, I would have predicted no recession for Canada. Unless you had told me there was a bad supply shock, but I don't think there was. But there was a recession.

Or, if in 2007 you had magically told me about the coming Canadian recession, I would not have predicted that CPI would follow those graphs. I would have predicted disinflation. But there wasn't any disinflation. Until very recently.

Sure, absolute downward nominal wage rigidity can prevent deflation. (OK that's not exactly true, because real wages could be rising, but set that minor caveat aside.) But it can't prevent disinflation. That 0% floor isn't always a binding constraint on all wages. Paul's graphs themselves show that.

Something is happening here. Or not happening. Or not happening until much later than I would have expected it to happen. And what it is ain't exactly clear. At least not to me.

Gotta go.

Comments

[Frances Woolley]

Nick or Stephen, can you unpublish this, make a minor edit, and then publish it again so that it goes above my post?

[Jim Sentance]

I commented on the pattern of inflation in the other thread but I think it got ignored, so I'll try again here and see what you think. During the period you're looking at there were fairly significant consumption tax moves by two of our bigger provinces (Ontario and BC) adopting the HST. Nova Scotia also raised their rate. At the time this was considered to have been a factor in raising the CPI above what it would have been. From what I've seen of the literature on how the HST works out, we would expect that over the following couple of years an offsetting fall in prices would occur as tax savings eventually get passed on by businesses. We're not talking huge shifts in CPI, but in the current context where the rate of change is quite low, I can't help but think this is big enough to contribute to the puzzle you see in the way the CPI has gone.

[Nick Rowe]

Frances: don't worry. It's fine.

Jim: there might be something to that. But I would have thought core CPI would have taken out the effect of changes in indirect taxes.

[marris]

Dammit it Nick! If you don't understand this stuff, then who does?!?!

Also, why aren't these questions enough to keep you from jumping on an NGDPLT (or any monetary policy) bandwagon? Do we really know more about NGDP than we know about the Philips Curve? If not, then doesn't that matter?

BTW, is there some list of nominal-real correlations that *haven't* broken down with the Great Recession? [I think Scott Sumner would say "sharp NGDP drop -> sharp employment drop" is on that list and has been observed since 1925].

[Ron Ronson]

Having read this blog for a while now the obvious answer is "because the Bank of Canada is not buying enough assets to keep inflation at 2%"

If wages are sticky and the demand for labor falls you would expect that an inflation target could be hit while RGDP falls (as happened in 2008 til now).

Harder to explain while the BoC's current interest rate target is failing to hit the inflation target even while RGDP is growing. Perhaps wages have become less sticky after 4 or 5 years of low demand and now you can get higher employment and RGDP growth combined with lower inflation for a while.

In both situations (sticky wages, low or negative RGDP growth) and (flexible wages, high RGDP growth) NGDPT would seem to be a better option that IT. Perhaps the BoC knows this and that's why inflation is being allows to fall or perhaps when disinflation sets in its not possible to hit an inflation target by interest rates alone.

[Mark]

Disinflation isn't happening b/c monetary policy is too tight. I see on CD Howe's MPC you agree :)

With nGDP growing on average 1.9% ann. over the last four quarters. Sumner has spoken!

BoC should stop targeting consumer debt levels imo.

[Stephen Gordon]

Maybe we did have a bad supply shock: the fall in the terms of trade is generally shoehorned into the standard AS-AD as an upward shift in the AS curve. It was maybe not so bad as to be the main driver of the recession, but perhaps bad enough to keep in mind.

[Lord]

Everyone believes in and follows BoC's lead in raising prices and lets output take care of itself, or more surreptitiously disguises cuts from them into the form of price discrimination, discounts, or promotions that don't make it into the stats or general sales. Inflation is the thermostat everyone has chosen to adjust to.

[Andy Harless]

How about the combination of these factors:

1. Some supply shock(s), not necessarily large enough to be obvious

2. Randomness

3. Serially correlated errors

4. Convex SRPC (implying it takes a long time to get a noticeable downward shift in the SRPC even when there's a major recession)

5. Downward sloping LRPC at low inflation rates (as implied by Krugman, and still relevant once you take into account 1-4)

I wonder also if the recent experience makes more sense when you assume sticky inflation, or maybe sticky information, rather than sticky prices.

[marcel]

IIRC correctly, Krugman at some point made the argument that with sticky wages, inflation is bound to remain slightly above zero: 2% is probably slightly (rather than well) above zero.

Suppose employers would like to cut wages of half their work force by 2% in real terms and raise the other half by 2%. The sticky wages (perhaps for Fehr reasons, or perhaps based on Bewlian reasoning) mean that they don't or cannot cut anyone's wages. Instead, half of their work force keeps the same nominal wage, the other half gets a 4% increase. The wage bill has increased by 2% (perhaps more if workers are shopping around for more money), inflation is 2%, and real wages are right where employers want them. That most but not all people are getting neither a raise nor a cut and some are getting a raise puts a positive floor beneath inflation.[1]

[1] Yes, I am ignoring productivity growth. You haven't mentioned productivity so neither (almost) do I.

[J.V. Dubois]

Nick, you are not the only one perplexed by this. A few weeks ago Tyler Cowen asked the same question but with even better example than Canada - Why do we actually see inflation in Greece? : http://marginalrevolution.com/marginalrevolution/2013/01/why-is-there-still-inflation-in-greece-model-this.html

And I remember that I asked you the very same question some time ago (how can there be an inflation AND simultaneously we can experience insufficient demand), and you answered that there seems to be a considerable inflation inertia caused by things like Calvo pricing etc.

Anyways I agree with you that none of the answers seems to be satisfactory ones - and for Greece even all of them put together seem very weak. So there seems to be something important here.

[Kathleen]

Are inflation expectations playing a role here?

I am always astounded at the reaction to any mention of CPI from Statistics Canada. There is an immediate flood of protests and single item examples supporting the assertion that inflation in Canada is much much higher than the published figures.

Are there any surveys of business attitudes towards inflation? The small business owners that I know (ok, not a random sample) all believe feverishly that inflation is huge because of the BOC low interest rates. If a business believes that inflation is present then there will be heightened price stickiness.

Perhaps price stickiness is not lower than wage stickiness?

[Paul Friesen]

Like Jim Sentance, I sort of feel like my previous explanation got ignored. It seems obvious to me that rapidly growing demand for commodities of many sorts from China is pushing up commodity prices, and that this is driving inflation in many parts of the world. Just look at oil prices for example. If you have rising commodity prices and downward wage rigidity, how would you generate deflation?

[Sam]

Recessions are still possible, even under a NGDP target. Once the nominal economic is pegged we must look to RBC theory. The US recession was a serious and very real shock on the Canadian economy.

[Nick Rowe]

There are two separate questions:

1. Is the BoC's monetary policy currently and recently too tight to keep inflation at the 2% target? Yes, I think it probably is.

2. But if the BoC's monetary policy is currently and recently too tight, why has unemployment continued to fall, and the economy continued to slowly recover?

This post is about question 2.

In other words, there's AD and AS. Bank of Canada policy is about AD. This post is about AS, specifically the Short Run AS, or the Phillips Curve, if you like, which is supposed to tell us how changes in AD get divided between changes in inflation and changes in output/employment/unemployment. There's something weird about the AS curve:

1. Either it shifted up/left (oil prices or taxes or some sort of supply shock).

2. Or it has become much flatter than it used to be.

3. Or there's a much longer lag between changes in employment and changes in inflation than there used to be.

My gut is leaning towards 3. If the disinflation we have seen over the last few months continues, that will confirm my gut feel.

[genauer]

The monetary policy of the BoC is to run an implicit peg to the US Dollar. Therefore the bond purchases, to match the yield curves.

That means, that all other goals, like exact 2.0% inflation, are secondary.

The Phillips curve might have described the UK from 1950 - 1960, or so, but as you can see with Greece or Spain, even with 25% unemployment, wages are barely moving, and so inflation.

The Phillips Curve, just like the Austrians of the 1930ties (from 1929 - 1933 wages did fall by 25% in the US), described some historic episode of more than 50 years ago, but not the reality today.

It is good this way. Sometimes we seem to learn at least a little bit.

And Greece seems to get at least a little bit of urgently needed deflation now. Their price level is about 10% too high, relative to real Europe. Milk is 60 cent the liter here, and the deviants will adjust their 1 Euro price to the median. And this is good for the average population there.

[Nick Rowe]

genauer: "The monetary policy of the BoC is to run an implicit peg to the US Dollar."

That is definitely false. During the last 20 years, while the BoC has kept inflation very near 2%, the US/CAN exchange rate has ranged from US$0.62 to US$1.05.

Stephen and now Scott Sumner are both arguing for a supply shock explanation, coming from a terms of trade shock. (Scott puts it slightly differently, but if your trading partner goes into recession, that is similar to an adverse terms of trade shock, I think). They may be right.

I'm still mulling this one over. My brain isn't really clear enough to give any useful responses to these worthwhile comments. Andy Harless gives a good list of possibilities.

[Nick Rowe]

Kathleen: "Are there any surveys of business attitudes towards inflation?"

The Bank of Canada does a quarterly Business Outlook Survey where it asks what they think inflation will be. Normally most expect between 2% and 3%, with a large minority expecting 1% to 2%. In the Winter survey 54% expect 1% to 2%, so expected inflation has fallen a little.

[edeast]

Aren't we losing the currency wars? Well until the recent fed misstep. We've had japan, us, and uk, all make big moves.

[Jim Sentance]

Nick, You're right of course that indirect tax changes are taken out of the core estimates, though I'm less certain that they would adjust for the process of eventually passing on the savings, so there might be some influence on the moderation end.

[Greg Ransom]

As James Carville ,ight say, It's the structure of production, stupid.

[Min]

What I have seen in the SF Bay Area, only in the past year, is a growth of discounts. Walgreen's has recently instituted a club, so that members can buy items that are marked down, temporarily on sale. Some pizza places are now advertising half price pizzas. Prices, at least at the low end, seem to be starting to give way.

[Kathleen]

"The Bank of Canada does a quarterly Business Outlook Survey where it asks what they think inflation will be."

This data is so interesting. Once I figure out how to download the tables I could see if there is any correlation between these various expectations e.g. forward looking inflation vs. price changes.

Wouldn't it take a down right expectation of deflation before prices would become unstuck? Expectations of a temporary decrease in sales volume could lead to reduced production runs and not a price reduction if a business expects stable, albeit low, inflation. A recent fall in inflation expectations could explain recent disinflation. Perhaps the credibility of the BoC has increased price stickiness?

I note that the sample size is small so any findings would not be entirely supportable I suppose.

[genauer]

Nick,

you were right until November 2007. In the moment the CAD overshot the USD to 1.07, this was corrected to the target 1.00 +/- epsilon within 2 weeks.
Same 2 weeks in April and July 2011.

I have been very verbose in my critic of Krugman and so many others from the American continent to give advice to Europe with very little knowledge and understanding, how it works. Soo, I am careful for myself into the other direction.

But this might be a rare case of you having your nose to close to the (Canadian) grindstone.

From the outside, the CAD peg to 1.00 is pretty obvious, including the instrument with which it is done (bond purchases of the BoC, starting mid 2011, every pickle in the exchange rate can be seen as a up/down kink in the balance sheet).

And it is the reasonable thing to do. It is attributed to Porfirio Diaz "Poor Mexico, so far from God and so close to the United States", having at least some nieghbors to the south. The same holds even more for Canada : - )

Furthermore, there are good reasons, why in Statistical Process Control (SPC), people are not permanently playing with the knobs, to keep this too exactly on target, and only act, if the measured value strays too far off the target.

[genauer]

I wondered, how to put this numerically, but simply.
Just draw up some histograms of various currencies in relation to the USD, over the last 1 years. The CAD is 0.998 (1.5% 1 sigma, 3 years 2.5%). 1.00 is not a random number.

The Hongkong Dollar at 7.77 (0.2% 1 sigma) neither. The Taiwan Dollar, Ruble at 30

I found some interesting links to many of those discussions:
http://eucenter.scrippscollege.edu/files/2011/06/06helleiner-cana.pdf
The Anglo Sphere being apparently VERY confident as early as 1942, to define the common global relations, Canada the only G-7 nation not going with fixed exchange rates after 1950.

Nick,
I understand your scepticism with regard to the peg much better with that

http://www.nber.org/chapters/c6867.pdf
Eichengreen on Bretton Woods / IMF history

[Ron Ronson]

"But if the BoC's monetary policy is currently and recently too tight, why has unemployment continued to fall, and the economy continued to slowly recover?"

Doesn't improved animal spirits (higher business confidence) explain it ? With neutral money RGDP would be surging ahead as the demand for labor (at current prices) increases. Tight money as the BoC fails (for whatever reason) to hit its inflation target slows it down a but not enough to prevent some RGDP growth.

[Spagianna]

Can I just say what a relief to come across someone who in fact knows what theyre talking about on the net. You definitely know methods to bring an problem to light and make it crucial. Additional men and women have to read this and understand this side of the story. I cant think youre not additional well known since you undoubtedly have the gift.


[url=http://cheapchinajordans.22web.org/]cheap jordans free shipping[/url]

http://louisvuittonhandbags8.fast-page.org/