« How much new revenue would be generated by an increase in federal corporate taxes? | Main | Generic Drug Pricing Reforms »


Feed You can follow this conversation by subscribing to the comment feed for this post.

Hmm, just several things:

1) Is it expectations of lower prices that drives AD down or is it expectations of falling future AD that drives down current AD and prices. There were countries with AD shortages even without any large impact on inflation. In such a case inflation is just unperfect signal, I think you once wrote about it like a dog that does not always bark when needed.

2) So CPI level seems OK - what about NGDP level? Is it back on trend too?

3) If NGDP is back on trend level, can there be slow recovery of unemployment partially due to structural reasons? Either some things that were there even before/during recession (like in UK) or there may be some new issues ranging from changes UI benefits to hysteresis effect of long slump. Is this what the whole PSST is about (I have to check)

4) Possibly even if inflation or AD is back on some new and lower trend for some time and economy did not recover, is it possible that there may be something inherently inferior in this new setup? Can there be long-run price of low inflation (or low AD growth)? Bryan Caplan has a nice article about this: http://econlog.econlib.org/archives/2012/09/long-run_unempl.html

Isn't it likely that the lack of momentum in recovery was initially due to poor household balance-sheets which do take time to recover?

Meanwhile, the sluggish recovery (at t) led to it creeping into future expectations of growth (ie at t+1) which pushed business investment down (which is based on NPVs - ie expectations at t+1 and beyond) leading to lower income growth for workers - not making deleveraging any easier.

JV: NGDP fell, started growing again, but isn't anywhere near back to trend. But the puzzle is inflation. Except for the initial big drop, CPI inflation has been above and below the 2% target during the recovery. Core inflation has only started to fall below target this last year.

Akshay: that is a story of why AD hasn't grown quickly enough. But if the Bank of Canada had wanted to make AD grow more quickly, it could have done so. It could have kept interest rates lower, and told people it was going to keep interest rates lower. It didn't do that, because it didn't want inflation to rise above target. Why didn't inflation fall, if it's purely an AD story?

(And the whole "people are repairing balance sheets" story sounds very strange to Canadian ears. Because a lot of Canadians seem to have been doing exactly the opposite, since personal debt/GDP ratios have been rising.)

Nick: Regarding inflation, I agree with you. I was just trying to explain that maybe price was not the primary mechanism through which this recovery could take place to begin with (at least initially) - debt principal reduction had to be the primary mechanism.

I also know that Canadian households are on a leveraging spree - but Canada's biggest trade partner has been on a deleveraging one (well, now the trend is reversing).

And the expected AD numbers for that particular trading partner seem to drive business investment decisions almost all across the world, not just Canada.

I know, inflation if puzzle. I was playing with FRED a little bit and what I find interesting is similarity to another country - Greece. Rapid disinflation during 2008 from 4% to zero followed by strong recovery to 5% in 2011 and then followed by rapid dive to 2% deflation now. However if one looks at inflation levels I do not know if anybody would say that Greece experiences probably the largest slump of modern history.

So if anything I am more in favor of your inflation stickyness theory. But I don't know.

Nick, I believe in the old-fashioned hypothesis that Pdot is a sign-preserving function of AD-AS. Hence, when AD > AS, commodity prices rise. When AD < AS, commodity prices fall. After all, prices are set by individual firms, not by omniscient "representative consumers".

Combined with your observations about price movements the recession was initiated by AD < AS but since 2009 we have AD > AS. Attributing slow growth to "lacking demand" may be a satisfactory explanation for businessmen who always fear "lacking demand". But a macroeconomists must attribute slow growth to other causes.

P.S.: You consider seasonally adjusted values of the CPI. Try the original CPI values! You will see very high quarter-to-quarter price movements which are at variance with any stickyness premise.

What bout debts?

Ugh, PSST. I can not recall Kling ever presenting any evidence for it over any other theory, he just kept saying it was a view he wanted to "push", generally without clarifying what he meant. I think it was Noah Smith who said PSST is a world where the invisible hand doesn't exist, and even normal economic times should be unable to deal with churn. Now actually looking over his post I see it wasn't him who brought up the point about churn (maybe Caplan did), but he did call PSST "the end of economics as we know it".
I also don't know what the point of "sustainable" is there, other than perhaps for Kling to say certain patterns he dislikes aren't sustainable.

Something like PSST makes sense to me intuitively. At least for small businesses, it's not unusual to rely on just a few big clients. If AD tanks and their business disappears, then you're done. I can imagine that over the whole economy this sort of dynamic can do a lot of damage that takes a long time to heal.



Its not as if builders get together to decide how many houses to build or supermarkets know with certainty how much Haagen Dazs they will sell next quarter. Its only a guess.

Like what Herbert suggested, if I've not misinterpreted what he meant, something like waves of AD and AS.

Also, Canadians levering up may be stalling or reversing. Hope I made some sense.

Sorry, What's a PSST?

Some examples in my minds of changing PSST:
70's oil shock: change in relative price of anything that use a lot of oil versus things that do not. You have liquidity in the newly-rich oil producers bank accounts, waiting to be spent ( in the long run the BP will be balanced, but in the meanwhile M effectively sustract from GDP not only from an accounting point of view but from the real one as well). We regear from producing consumption goods for the western consumers to building massive amoiunts of arms for oil sheikdoms. Question: should monetary ploicy have been accomadating to replace the temporarily dormant liquidity stashed in swiss bank accounts instead of being restrictive to "fight inflation" and thus not only prolonging the "recession" and hampering the restructuring process by making borrowing for investment more expensive?

Québec restructuring: from the late 50's to the early 00's, unemployment seemed high in QC and in fact was even worse than official rate ( see the Excel spreadsheet I posted a couple of years ago). Cdn-anglo press full of "political instability caused by bad separatists etc". " Now GDP/cap no longer at 60% of Canada (itself growing and changing composition)and U in the ON range. Would a separate monetary ploicy (aka own currency?) have facilitated the process?

Greece: conservative german and u.s. press full of "Greeks retire at 50-55-whateverseemscandalous no wonder they're in trouble!". And if it was restructuring? Putting old guy to pasture so investment is directed to better use employing younger and better trained people? And politically, an old guy with a pension is way less likely to throw rocks than an unemployed young guy who sees his future melting (no job,no money, no car, no girlfriend,no house, no children). As I tell my students, sociologists and police officers have a name for an unemployed 15-30 year old male : TROUBLE"...

Wonks: The way I see it, the invisible hand works, but not instantly like a magic wand. It takes real people (some call them "entrepreneurs") real time and resources to find profitable opportunities, and bring everyone together. Coordination doesn't just happen instantly. Yes, there is always churn, and economies deal with it, but not instantly and perfectly. But an AD shock (which is where my hybrid view differs from the standard PSST view, is like a very big real shock, if prices are sticky.

I read "sustainable" as: "Hey, let's try this....waits a few months or years....Hey! It works! We're making a profit!"

What if restructuring can't occur without inflation? Entrepreneurs don't hire without demand. They may innovate for cost savings but not for novelty, or may under invest without the safety net of inflation and increasing demand.

Are there limits to growth when deleveraging? Canadians are doing well on net worth so they can afford more debt. Who was it that did that fine analysis of historical sustainable debt burdens and cyclicality?

Interesting post and comments.

One thing I'm confused about: do you think the Canadian economy has not yet receovered from the recent recession? Unemployment is near NAIRU for Canada, employment is rising with trend (with the new trend that is although perhaps there was a permanent loss relative to the pre-crisis trend), and inflation is positive along with GDP growth. Seems like a relatively solid recovery.

What is there to be explained? Why the recovery isn't faster? I thought the general model is that growth/recovery is when the economy is operating normally and the boom and recession times are when something is unusual requiring concepts like AD shortfalls.

Looking forward to understanding better.


Enon: my sense is that the Canadian economy has not yet fully recovered. Inflation is below target and falling. Unemployment is still a bit high. Output below trend. One never knows for sure.

To be explained: why wasn't recovery faster, and why was there at or above target inflation during most of the recovery.

Nick, thanks for the reply. That clarifies things.

A couple of more questions:

1. How would you define fully recovered? According to the OECD (http://stats.oecd.org/Index.aspx?QueryId=36376), NAIRU in Canada was 7.4% yet unemployment is 6.9% (http://www.cbc.ca/news/business/canada-adds-22-000-jobs-in-november-1.2453441). What measure would you use to define recovered? I'm not trying to be argumentative here. I think you could easily make the point that Canada is not recovered and the NAIRU is wrong (either as a concept or its estimate by the OECD) since inflation is currently low and falling. Perhaps a GDP output gap? The trouble with those is they can be sensitive to trend estimates, revisions, estimates of potential growth and so on.
2. Why do you say "why was there at or above target inflation during most of the recovery"? Do you mean *below* target? Data from (http://www.rateinflation.com/inflation-rate/canada-historical-inflation-rate) suggests inflation was below 3% from 2009 to present except for a couple of months.


NAIRU as a concept? I remember someone I respect ( I don't remember which one exactly, don't want to make a mistake...) saying that NAIRU is the average ( algebraic,geometric, logarithmic, pick anything) of the last (3,5.7, pick one) years. What he meant was that if you push AD enough, people will get employed because someone will find something to make because someone will want something to buy. Wrongly managed it might mean an unsustainable bubble but it may not be compulsory.
The current ECB policies means that NAITU will go up,they will call it "structural", call for "reform" (code for lower wages and higher take for the 1%).
NAIRU is the modern incarnation of the lump of labor fallacy.

I'm OK with NAIRU (aka the natural rate) as a concept, with the understanding that it is only a well-defined concept in some models. But empirical estimates of NAIRU (and the output gap), especially real-time estimates, are very unreliable. Simon van Norden, for example, has found that the Bank of Canada's real-time estimates of the output gap have little resemblance with estimates made with hindsight. And if we do take the definition of NAIRU literally, inflation has been falling this last few months, therefore unemployment must be above NAIRU. But inflation was rising earlier in the recovery, so the unemployment then, even though higher than today, must have been below NAIRU. You could say that what I am doing in this post is redefining a dynamic version of short-run NAIRU.

Plus, I don't think booms and recessions are symmetric, so I do not think that unemployment will equal NAIRU on average when there are AD fluctuations. Yet something equivalent to that assumption is imposed on empirical estimates of NAIRU.

The Bank's inflation target is 2%. Given that unemployment was a lot higher than it had previously been, why wasn't inflation well below 2%?

Jacques: "NAIRU is the modern incarnation of the [that which must not be mentioned for fear of bringing in a particular commenter] fallacy."

I have always thought it was the opposite of the LOL fallacy. In it's crudest (and misunderstood) version, it says that nothing causes unemployment to rise.

Nick: in its crudest form, beloved in some quarters,it also says that nothing could cause it to get lower, at leat in the medium-to-long-term. And that nothing should be done to get it lower. Very useful if you want to maintain the reserve army of the unemployed. Or justify your ECB -EC policies.
Or maybe that's just me.

All those are good points.

I personally don't have a favorite measure of when we are in recession vs recovery vs boom, etc. (but I would like to!). Hence the question: how does one actually define when the economy is recovering slowly vs where it should go? Seems like a basic question. Probably the answer won't be perfect (e.g., the issues with NAIRU or output gaps). But without at least some rough measure, it seems hard to discuss the question of why the Canadian economy has been recovering slowly. (By the way, I think a reasonable answer for the slow recovery is that the CAD has been relatively strong due to most other major economies pursuing record monetary easing in a variety of forms. Of course, this is not the easiest thing to put into a model either...)


I think you answered your own question in the comment starting “Wonks. The way I see it…”. What you refer to there is Say’s Law isn't it? And I realise that Say’s Law doesn’t work as well in a complex economy as in a simple barter economy, but I don’t think it’s necessarily completely stymied in complex economies. I’ll expand on that, in the unlikely event that anyone is interested in my quaint ideas about Say.
Unemployed or underemployed entrepreneur and employees get together and produce commodity X and offer it for sale at slightly below the prevailing price. That way, they ease their way into the existing economy, and the price of X falls marginally compared to other goods.
Then a bunch of people do the same with commodity Y, Z etc etc, etc.
That seems to suggest that there is a fall in prices in the aggregate. But that’s not necessarily true. E.g. if there are underlying forces making for a bit of constant inflation, that doesn’t destroy the above way in which producers of X, Y, etc ease their way into the economy. I.e. as long as there is a bit of price flexibility at the micro economic level, then Say’s Law works.
In contrast, if households try to hoard cash (Keynes’s paradox of thrift), then Say’s Law is stymied. So we shouldn’t rely on Say’s Law: i.e. standard methods of bringing stimulus are preferable.
Re AS and AD lines, charts, etc., I just don’t buy those. The important point (as you suggest) is simply the desire by individual entrepreneurs and potential employees do something. And as long as there’s a bit of price flexibility at the micro economic level, that will work. Prices in the aggregate are unimportant.

Enon: a high exchange rate doesn't really help explain anything here. A high exchange rate would reduce AD, other things equal. But if the Bank of Canada had wanted to, it could have increased AD.

I don't have any brilliant insights into how to estimate full recovery. If it is abnormally easy to buy things with money, and abnormally hard to sell things for money, I say there is a shortage of AD, and the economy is in recession. But "abnormal" is a loose word. And there is a distinction between levels and rates of change. Hiring more labour may be easy; but hiring more labour more quickly than you are currently hiring may be hard. It is that dynamic question that this post is about.

Ralph: "I.e. as long as there is a bit of price flexibility at the micro economic level, then Say’s Law works."

That's where I would disagree. Fallacy of composition, and all that. An individual producer can always sell more by cutting his price. But it doesn't necessarily mean that all producers can sell more by cutting all prices. They might just be taking demand away from each other. The slope of the AD curve matters.


I’m aware of the dangers of falling for fallacy of composition. If I’ve made a mistake, that’s not it. What I’m saying, to put it in different words, is this.

Start with a barter economy. Robinson Crusoe lands on desert island where a few families live. Initially he is unemployed. But his speciality is fishing, so he does that, and the price of fish drops in terms of other goods and services. That microeconomic change in prices enables Crusoe to merge with the existing economy.

Next, the economy introduces money. That makes no difference: as long as the price of fish drops relative to the price of other goods, Crusoe merges with the existing economy.

That process can be repeated for others landing on the island and who specialise in producing other goods (grass skirts, mud huts, etc). In each case, a MICROECONOMIC change in prices is needed to merge each newcomer with the existing economy.

The comments to this entry are closed.

Search this site

  • Google

Blog powered by Typepad