Japan data

Japan bashing is popular. Looking at the data since QE started:
Japan-GDP-per-capita--Constant-Prices-Since-2000--Chart-000002

Real per capita GDP has risen 1.4% a year and yet prices have been constant.

I’m not sure why any one inflation target (positive or negative) is better than any other ((Woodford chapter 7 argues a target that tends to zero is optimal or a target that has expectations of inflation tending to zero is optimal.)). In any case, the level of the target is of second order importance to having stable inflation (and thus stable inflation expectations and thus having a target). The costs of discretion are much larger than the costs of minimal inflation (or deflation).

6 thoughts on “Japan data”

  1. To be sure, Japan barely ever did any QE — $300 billion worth, after which they did have growth (probably b/c exports picked up — nothing to do w/ the QE), after which they (I think foolishly) reversed all of their QE.

    The Japan experience is not informative in any way about the efficacy of QE.

  2. Will,

    I am a bit confused by your comments here.

    Suppose that you have a target rate of inflation of 2%. Now suppose that the price level drops by (I’m just making this up) 3%. Do you continue to have a 2% target for inflation the next year? This is the question that Sumner has been asking.

    The problem is that Japan has (basically) a 0% target, but then when AD falls, they seemingly allow the price level to fall and do not adjust their inflation target going forward. Given that they target 0% inflation, they are not targeting a lower price level. By adopting a price level target, they would attempt to get the price level back to its previous trajectory — trajectory is likely the wrong word given that they are targeting 0% inflation.

    I am not much for the price level or inflation targeting, I prefer a nominal income target. I think that it is more intuitive and I think that it fits better with monetary theory. However, the same rationale essentially holds.

  3. EE, I agree history dependent policy (e.g. level targets or a state-contingent inflation target) may be better than a fixed target. This recession with rates going to zero and all the silly liquidity trap talk is probably proof of that… BUT its not a good idea to change policies mid-recession. Doing so runs the risk of appearing to have discretionary policy which would make expectations much harder to manage. This outcome would make things much, much worse for much longer.

  4. I think it depends on how well you articulate what you are doing. For example, if the Fed were to adopt Svensson’s “Foolproof Way”, they would be required to set their price level target and their “exit strategy.”

    In other words, I am inclined to agree with you provided that the Fed came out and said “we are now targeting X% inflation” and provided no guideline for the future. However, if they provided some guidance as to how policy would be conducted in the future, I don’t think that the cost of discretion would be very high because they are essentially outlining future policy according to a rule.

  5. I just don’t think we know enough about expectations setting for what you say to be ok. I don’t think speeches from Ben makes for credible policy.

    Expectations dynamics would be interesting area to study.

  6. The Fed currently operates under pure discretion. Just because we can roughly describe their behavior with a Taylor rule, does not mean that they follow such a rule. As Taylor’s own work shows there are long periods of time in which the Fed has deviated from the rule that otherwise provides a reasonable description of their behavior.

    If Bernanke comes out and says that he is going to target the price level and he isn’t seen as credible, how is this different from what we have now. I turn on financial news networks and I see people debating what the Fed is going to do next and whether we are headed for hyperinflation. I don’t see how outlining a specific policy makes our situation worse.

Comments are closed.