I’m puzzled as to why economists are puzzled

Here’s an interesting WSJ report on Canada:

And the Bank of Canada has signaled it’s aware of the downside risk to inflation prices in light of the long string of below-target inflation numbers.

The broad weakness in price pressure across a number of economists is hard to account for, Roberto Cardarelli, chief of the IMF team covering Canada, said a press briefing Wednesday.

“Inflation is a puzzle, and it’s not just a puzzle in Canada,” Mr. Cardarelli said.

It’s believed that one of the key drivers of inflation is the amount of slack in an economy – the difference between actual growth and the growth an economy could produce without pushing inflation higher.

It is also referred to as the “output gap,” as it’s the difference between an economy’s actual output and its potential output. When an economy is reaching full capacity, labor and other inputs become scarcer and more expensive, and that creates inflation.

But inflation in Canada remains soft despite the fact that the gap has narrowed considerably since the recession, Mr. Cardarelli said.

What’s the big puzzle?  In the textbook AS/AD model, when AD shifts left output falls and the price level (or inflation) also declines. Then the self-correcting mechanism kicks in.  AS shifts right, causing prices (or inflation) to fall further and output gradually rises back to the natural rate.  As long as NGDP is stable, the rising output will be associated with low inflation.

It seems like lots of economists abandoned the excellent AS/AD model and replaced it with a very unreliable output gap model.  If they had studied the Great Depression they would have realized that NGDP growth drives inflation, not output gaps.  In 1933 NGDP growth was very rapid but unemployment was 25%.  The WPI rose by more than 20% in the 12 months after March 1933, which confirms the NGDP model, but is completely inconsistent with the output gap model.

HT:  Clare Zempel


Tags:

 
 
 

9 Responses to “I’m puzzled as to why economists are puzzled”

  1. Gravatar of W. Peden W. Peden
    2. December 2013 at 11:41

    “It seems like lots of economists abandoned the excellent AS/AD model and replaced it with a very unreliable output gap model. If they had studied the Great Depression they would have realized that NGDP growth drives inflation, not output gaps. In 1933 NGDP growth was very rapid but unemployment was 25%. The WPI rose by more than 20% in the 12 months after March 1933, which confirms the NGDP model, but is completely inconsistent with the output gap model.”

    Would I be right in thinking that this is a dispute over the flatness and/or predictability of the SRAS?

  2. Gravatar of ssumner ssumner
    2. December 2013 at 11:50

    W. Peden, Not just the flatness, as a shift to the right in SRAS is deflationary for any SRAS curve that is not 100% flat. And I don’t think anyone believes it’s 100% flat.

  3. Gravatar of dtoh dtoh
    2. December 2013 at 11:52

    Scott,
    Ignoring the models, my experience is that producers will raise prices when there is an increase in the expected rate of RGDP growth causing an expected short term imbalance in supply and demand. Even with high employment, producers are limited in their ability to increase production over the short run.

    In general, it’s difficult to increase production over the short run because of a lack of manufacturing capacity (after an extended recession, idle capacity is usually fairly low) and because even with the ability to hire workers, it takes a fair amount of time to actually train employees before they become productive.

  4. Gravatar of ssumner ssumner
    2. December 2013 at 14:20

    dtoh, Yes, those are just part of the problem with output gap models.

  5. Gravatar of Lorenzo from Oz Lorenzo from Oz
    2. December 2013 at 16:42

    “I don’t think anyone believes it’s 100% flat.” Not if asked explicitly, but if one expects no change in P (or no change in the rate of change of P) until some (short run) response limit to changes in y is reached, then what else is one saying?

  6. Gravatar of Lorenzo from Oz Lorenzo from Oz
    2. December 2013 at 16:43

    Which would make the output gap model a way of hiding from oneself assumptions one would not explicitly admit to. Not a good thing in a model.

  7. Gravatar of Doug M Doug M
    2. December 2013 at 19:03

    It always seemed to me that you NGDPLT model was a variation off of an output gap model. In the output gap model, you take the previous high level of RGDP, you then slap on a constant growth rate of population and productivity. To the extent that RGDP is below that projection there is “slack.” How different is that from NGDPLT, with the exception that you have replace RDGP + C to NGDP + 5%?

    AD/AS models stink, because neither AD nor AS can be observed. What we can see are changes in output, sales, prices, and employment.

  8. Gravatar of Daniel Daniel
    3. December 2013 at 01:15

    Scott, if output gaps are mostly misleading, does that mean that we should prefer NGDPLT targets to stay unchanged, even when huge disasters happen?

    Toy example: war devastates Examplestan. Examplestan ends the year with both NGDP and RGDP suddenly cut in half.

    Should the Examplestan central bank aim to get the country back on the unchanged prewar path for NGDP?

    Are there special considerations to getting back on track with NGDPLT in the aftermath of a national disaster? Or is that just like any other deviation from target, only bigger?

  9. Gravatar of ssumner ssumner
    3. December 2013 at 11:12

    Lorenzo, Good question. I don’t know the answer. I suppose they have some dynamic story.

    Doug, I consider AD to be NGDP, which is observable. Yes, the SRAS is unobservable, but so are output gaps.

    Daniel, I actually prefer targeting NGDP/person, but often leave that part out because in the US the pop. grows at a steady 1% per year. The NGDP/person target should not change after a disaster.

Leave a Reply