A worthwhile Canadian initiative

Over at WCI, the Bank of Canada has put out a request for monetary policy reform ideas.  This is from a post by Carolyn Wilkins, of the BOC:

During the Great Depression, John Maynard Keynes wrote “It’s astonishing what foolish things one can temporarily believe if one thinks too long alone, particularly in economics.”

Today, this is truer than ever, and it’s a principle the Bank of Canada is keeping top of mind as we tackle some tough policy and research questions. As I explained in a speech last week, we’ve been thinking a lot lately about “innovation, central-bank style.” That means taking on board the lessons from the most recent global financial crisis, questioning old answers to policy questions and changing our corporate culture to remove obstacles to new ideas.

It starts with a clear research agenda. Here are some highlights:

  • While we’re committed to our inflation-targeting framework, we’re taking a hard look at it ahead of the official renewal in 2016. The questions we are asking are: Should we consider targeting a rate of inflation higher than 2 per cent? Should we continue to use CPIX as our operational guide? How do we incorporate financial stability risks into the conduct of monetary policy?
  • We’re analyzing the effectiveness of unconventional monetary policy used by various countries since the crisis. We’ve just published staff discussion papers on quantitative easing, forward guidance and negative interest rates.
  • We’re making a big investment in economic modelling and using new analytical methods like big data and behavioural economics.
  • We’re preparing to confront “alternative futures” such as China’s new role in the international monetary system; new payment methods like Apple Pay and PayPal; disruptive technologies and business models like Bitcoin, peer-to-peer lending and others in the sharing economy.

To get to the right answers to these questions, we need to reach out to people beyond our own four walls. If we don’t, we leave ourselves open to confirmation bias and tunnel vision—and we risk missing out on insights that can ultimately lead to better outcomes for Canadians.

That’s why I’m reaching out to you—those who follow and contribute to Worthwhile Canadian Initiative.

Since they are committed to inflation targeting, I’m not going to recommend NGDPLT. In any case, NGDP targeting may not be optimal for Canada, given the big swings in nominal commodity output.  Instead it might be best to stabilize the path of nominal total labor compensation.  So here I’ll advocate a form of flexible inflation targeting.

Each year, the BOC will estimate the rate of growth in total nominal labor compensation (TNLC) that is consistent, in the long run, with 2% inflation.  Thus suppose they forecast a trend rate of 1.2% labor force growth and 0.9% productivity growth.  In that case, 2% long run inflation is consistent with 4.1% growth in TNLC. Then 4.1% becomes this year’s TNLC target.

Importantly, I am considering a level-targeting regime, although the BOC could adopt a growth rate targeting approach if level targeting made them too queasy. Suppose that one year later, the BOC forecast 1.1% labor force growth, and 0.8% productivity growth. Then this year’s target would have TNLC rise by 8.0% over the level of 2 years ago (4.1% for last year plus 3.9% for this year (ignoring compounding).)  That’s if they choose the level targeting approach.  If they go with a “let bygones be bygones” growth rate approach, then they’d aim for 3.9% compensation growth during this year, from the point where TNLC left off at the end of the previous year.

This regime would assure roughly 2% inflation in the long run, while allowing some year-to-year fluctuation to help stabilize the labor market when the economy is hit by real shocks.  If BOC forecasts were unbiased, then inflation fluctuations should be fairly random. Actual inflation would become somewhat countercyclical, which is the biggest difference between this plan and current central banking policy.  Recall that in most countries inflation fell during the severe 2009 slump.  Under my proposal it would have risen.

PS.  If they insist on staying with simple inflation targeting, I’d raise the target range to 2% to 3%, as in Australia.  Recall that the Aussies easily avoided the liquidity trap in 2009. The zero bound is much more likely to occur in the future than in past recessions.  However under my proposal described above, especially the level targeting approach, no increase in the inflation target is needed.



16 Responses to “A worthwhile Canadian initiative”

  1. Gravatar of Benjamin Cole Benjamin Cole
    20. November 2015 at 08:28

    You know, the IT band is sort of the poor man’s NGDPLT. It may be a good substitute. I happen to think a 2.5% – 3.5% IT band for the US would work good enough.

    It is distressing to ponder that much of what is preventing full-throttle economic growth is a small difference in IT targets.

  2. Gravatar of Kevin Erdmann Kevin Erdmann
    20. November 2015 at 08:46

    In level targeting, is there a risk of economic dislocation if something causes a positive jump in GDP, wages, or inflation, especially if that jump is specific to limited sectors? In that context, could level targeting become destabilizing because it would create negative price and wage pressures?

  3. Gravatar of ssumner ssumner
    20. November 2015 at 10:09

    Kevin, If it were level targeting of prices, I’d say yes. But that’s also a risk with IT.

    If it’s level targeting of labor comp, as I propose, then I think the risk is very low, due to sticky wages. And I’d add that growth rate targeting would be even more destabilizing.

  4. Gravatar of LK Beland LK Beland
    20. November 2015 at 11:36

    What lessons can we take from South Korea? They seem to have had no true recession in the past 13 years. Not a bad run at all, in my opinion.

  5. Gravatar of Jerry Brown Jerry Brown
    20. November 2015 at 12:45

    Professor, this is very interesting. I think I like total nominal labor compensation level targeting even better than NGDPLT targeting. But I have a few questions.

    If I understand it correctly (Big If there), TNLC would tend to make sure that real labor compensation would rise (or fall) with productivity increases (or decreases). Is that right?

    And would TNLC be more countercyclical in nature than NGDP targeting or about the same? And would you also recommend it for the U.S.?

  6. Gravatar of Jerry Brown Jerry Brown
    20. November 2015 at 13:05

    While I’m pestering you with questions, I have one more. Do you think that over time, if estimates of TNLC proved accurate, and if the Central Bank was able to target it effectively, would the Central Bank have to continue to keep a 2% base inflation rate as a target, or could they lower that safely?

  7. Gravatar of TravisV TravisV
    20. November 2015 at 13:38

    Why Is Janet Yellen Panicked About The House’s FORM Act?


  8. Gravatar of Ray Lopez Ray Lopez
    20. November 2015 at 20:23

    This reminds me of a quote by Friedman, in the late 60s or 70s, that says something like everybody has a different rule of thumb for monetary policy, and everybody thinks their rule is best.

    It’s interesting that Sumner has different rules for different countries: NGDPLT for the US, IT for Canada. And for an open country that trades a lot? Should they just adopt the dollar, like Costa Rica does?

  9. Gravatar of James Alexander James Alexander
    21. November 2015 at 07:19

    Draghi mentioned a problem with the “velocity” of bank reserves. Perhaps the start of some really fresh, monetary, thinking?

  10. Gravatar of TravisV TravisV
    21. November 2015 at 11:53

    Bonnie Carr with a great find:


    “Fed’s Williams says central banks need to consider new tools”

  11. Gravatar of TravisV TravisV
    21. November 2015 at 13:21

    ECB executive board member Benoit Coeure: Monetary policy risks becoming ineffective in low growth world


  12. Gravatar of Major.Freedom Major.Freedom
    21. November 2015 at 15:10

    Carolyn Wilkins, yet another dime a dozen Keynesian who has no clue that it is central banking itself that was responsible for the financial crisis.

  13. Gravatar of Benjamin Cole Benjamin Cole
    21. November 2015 at 16:34

    Another reason and IT band might work, is that central bankers try to forecast inflation. The whole long and variable lags….

    So if you have an IT ceiling or a target, central bankers tend to undershoot. But if central bankers have a IT band, they may feel they have a little more leeway before suffocating the economy.

  14. Gravatar of ssumner ssumner
    22. November 2015 at 09:24

    LK, They probably have good monetary policy.

    Governor Brown, Monetary policy has no long run effect on real wages. Yes, in the short run real wages would tend to rise and fall with productivity, as they should. For the US, either policy would be fine, the difference would be minimal.

    I would hope that at some point they abandon inflation entirely, and just target NGDP/person. Inflation doesn’t matter.

    Ray, No, I don’t favor IT for Canada.

    James, Let’s hope so.

    Travis, Thanks, check out my Econlog post.

  15. Gravatar of LK Beland LK Beland
    22. November 2015 at 10:20

    Me: “What lessons can we take from South Korea? They seem to have had no true recession in the past 13 years. Not a bad run at all, in my opinion.”

    Prof. Sumner: “LK, They probably have good monetary policy.”

    I tend to agree. It is a rare thing in the developed world. At the moment, their rate has inched down to 1.5%. They are now in a situation where most other central banks have failed, except maybe for Switzerland (its labor market seems OK). What to expect now?

    I would certainly appreciate a post about this issue, at some point in the future.

  16. Gravatar of ssumner ssumner
    23. November 2015 at 18:13

    LK, They differ from other developed countries in that they were not fully developed in 2008. They had a higher trend rate of growth, and that’s one thing that kept them above the zero bound.

Leave a Reply