Sunday, February 23, 2014

Japanese inflation isn't as high as you think



I always thought that "core-core" was a kind of hardcore music so hardcore that the only one who can play it is GG Allin, and only if he literally blows himself up onstage. But actually, "core-core" is a type of Japanese inflation that does not include food or energy prices. You may recognize this as being the same thing as what the U.S. calls "core" inflation. And you'd be right. The problem is that Japan already had something that they called "core" inflation, which omits food but does include energy prices. This naturally produces confusion in the press, since journalists dutifully report on Japanese "core" inflation, which readers take to be the same as the U.S. measure, even though it isn't.

Anyway, so why is this important? Because for months now, you've been hearing that Abenomics - or at least the monetary policy part of Abenomics - has been a solid success. Japanese inflation, you've heard, is climbing up toward the 2% target. To many people, that is a sign that monetary easing can hit inflation targets if the central bank is really committed to doing so.

The problem is, the Japanese inflation that people are looking at is the Japanese "core" rate, not the "core-core" rate. In other words, those rosy numbers you're seeing include energy prices. And energy prices have gone up, partly because of supply restrictions that are unique to Japan (i.e. restrictions on nuclear power in the wake of Fukushima).

Sober Look has the details:
The upturn in [Japanese "core"] CPI inflation by December was still being heavily influenced by utility prices that were up 5.5% y/y due to the effects of yen depreciation on imported natural gas and oil prices, and higher electricity prices in the face of Japan’s continued shutdown of all of its nuclear reactors. Take energy and food — which is also probably under upward pressure in part due to yen depreciation — out of CPI and it is only up 0.7% y/y as food prices themselves are up 2.2%. Most of the CPI effects of the Bank of Japan’s efforts to depreciate the yen remain confined to a relative price shock to food and energy...
 
In other words, if you use the inflation measure we use in the U.S., Japanese inflation is running at only 0.7% - not very close to its 2% target.

And as Paul Krugman reminds us, energy prices are volatile, so we shouldn't think about them when we assess the short-term effects of monetary policy:
[O]fficial Fed doctrine is to focus on core inflation, not react to short-run fluctuations in commodity prices. And the history of the past decade or so has showed that this is very much the right thing to do — headline inflation has swung widely, while focusing on core inflation has been a much better (though not perfect) guide to appropriate policy: 

(Now that's PCE, not CPI, but the point is the same.)

So basically, Abenomics has not yet shown that a central bank can hit a 2% inflation target after a long period of deflation. That proposition remains an article of faith. Perhaps the target will be hit...perhaps not. (Of course, if it's not hit, expect a few supporters of monetary easing to say that the Bank of Japan was just not committed enough to hitting it...)


Update: I recently bought a Japanese-language DVD biopic of GG Allin in a back-alley shop in Harajuku. I have now discovered the true meaning of "core-core".

23 comments:

  1. I make some estimates of the impact based on some counterfactuals here (using "core core" CPI):

    Link

    It really is hard to say much is happening at this point. It's pretty obscured by empirical and model error.

    ReplyDelete
    Replies
    1. Anonymous7:35 AM

      It really is hard to say much is happening at this point. It's pretty obscured by empirical and model error.

      This would not stop a serious economist from making sweeping judgments.

      Delete
    2. Here's an update RE Scott Sumner's claim of evidence in favor of inflation ... there's no monetary effect

      Delete
  2. Anonymous9:15 PM

    The Important inflation is wage inflation. With slack demand for labor and a global labor market, good luck meeting any inflation target with monetary policy. Regulatory and fiscal policy could have much greater effect on wage inflation. Raise the MinWage. That is one regulation that would help the Fed meet its target. Bring projects and repairs forward, do them now when there is excess labor. That would strengthen the labor marker and give labor a little better bargaining power. We are not using the right tools to get wages moving upward.
    -jonny bakho

    ReplyDelete
  3. "Of course, if it's not hit, expect supporters of monetary easing to say that the Bank of Japan was just not committed enough to hitting it..."

    Well, if they did it 5-fold, that wouldn't cause at least 2% inflation? 20-fold? 100-fold? It's a lot to argue that quantitative easing has no effect on inflation, because you're arguing 100-fold would do nothing -- at all -- markets are that perfect, and the population is that perfectly rational, expert, and informed.

    ReplyDelete
    Replies
    1. Alternatively, you're just using a different framework, one where QE is just an asset swap and does not create any sort of 'hot potato' effect.

      Delete
    2. Anonymous5:48 AM

      That's nice dear

      Delete
  4. "So basically, Abenomics has not yet shown that a central bank can hit a 2% inflation target after a long period of deflation. "

    It largely depends on who the counterparties to the central bank are in MP operations. If you deal with entities that have little or no marginal propensity to consume then expanding money to them wont be very effective in reaching the inflation target.

    ReplyDelete
  5. The BoJ is getting much closer to hit the target on the price index which the BoJ is targeting. But this is not evidence that the BoJ can in fact hit an inflation target... because there exists another price index which the BoJ isn't targeting, and BoJ hasn't hit a 2% rate of change on the price index the BoJ isn't targeting.

    Nice argument. Are there any price indices which the BoJ isn't targeting, and the BoJ is failing to hit a 2% rate of change on? I think we should be told.

    ReplyDelete
    Replies
    1. argosyjones10:03 AM

      Noah's point is that because energy prices are volatile, energy supply shocks, instead of monetary policy, are driving 'japanese core' inflation. I think this implies that when additional energy supply comes online, or demand diminishes, prices will decelerate or even begin to fall, and the target will no longer be met.

      That seems pretty persuasive to me.

      Delete
    2. Anonymous3:58 PM

      Also, while the BoJ is free to set a policy of aiming for a target that is largely beyond its ability to control, those of us who are interested in judging whether it's a good policy might do well to focus on a target that it's within the bank's control to hit or not.

      Delete
  6. This is the sort of pandemonium and chaos you get when we stop teaching the Marshall Lerner thing in schools...

    http://www.youtube.com/watch?v=JmzuRXLzqKk

    ReplyDelete
  7. "...But actually, "core-core" is a type of Japanese inflation that does not include food or energy prices. You may recognize this as being the same thing as what the U.S. calls "core" inflation. And you'd be right. The problem is that Japan already had something that they called "core" inflation, which omits food but does include energy prices. This naturally produces confusion in the press, since journalists dutifully report on Japanese "core" inflation, which readers take to be the same as the U.S. measure, even though it isn't...The problem is, the Japanese inflation that people are looking at is the Japanese "core" rate, not the "core-core" rate. In other words, those rosy numbers you're seeing include energy prices...In other words, if you use the inflation measure we use in the U.S., Japanese inflation is running at only 0.7% - not very close to its 2% target..."

    Not only does this produce confusion among journalists, it evidently elicits a lot of confusion in a certain economics blogger as well.

    Only three weeks ago in a blog post entitled "What can Abenomics teach us about macro (so far)?" you stated

    "Headline and core inflation are both up, now into positive territory. Core inflation is higher than it has been in decades"

    And linked to the following graph:

    http://1.bp.blogspot.com/-96gs1McYifE/Uu5xwPzEq7I/AAAAAAAABXo/HHz6Nt6E5Dk/s1600/Japan1b.png

    The graph shows CPI and core-core CPI although core-core CPI was mislabeled as "core". But there was no harm, no foul, because "core-core CPI" is essentially what the rest of the advanced world calls "core CPI". Moreover you were right when you said it was the highest in decades because year-on-year core-core CPI last exceeded 0.7% in August 1998, and only then really because of a 2 point increase in the consumption tax starting in April 1997. Prior to that one has to go back to March 1995 to find a year-on-year core-core CPI rate of inflation that is higher.

    So the main revelation of this post appears to be that you didn't know what you were posting the last time you discussed Japanese inflation, and that now you don't know that you got it mostly right then despite your ignorance.

    P.S. Didn't you once imply that you know more about Japan than the average economics blogger?

    ReplyDelete
    Replies
    1. Bill Ellis6:01 PM

      Mark gets catty ! Reeooow. Hisssss.

      The point is that using the "Core" instead of the "Core Core" misrepresents the effectiveness of BoJ monetary policy. And that is bad news for Monetarists.

      Delete
    2. Who is using core? Three weeks ago Noah Smith posted a graph of core-core even if *he* didn't know it. Secondly, how is it bad news that core-core is the highest rate it has been in nearly two decades?

      Delete
    3. So the main revelation of this post appears to be that you didn't know what you were posting the last time you discussed Japanese inflation, and that now you don't know that you got it mostly right then despite your ignorance.

      Possibly. Or possibly I knew all about the different definitions of "core", and used the word "core" to refer to "core-core" because I knew how readers would understand the word. ;-)

      Delete
  8. Noah,

    Something about Japanese inflation doesn't make sense to me. The yen has depreciated rapidly over the last year and a half, but the CPI has barely budged. (I think I'm looking at the non-core CPI, which was 100.9 in December.)

    If higher prices for imports are mainly due to depreciation of the yen, then prices of domestic goods must fall in lockstep to offset the depreciation of the yen. Why would this occur?

    ReplyDelete
    Replies
    1. Frank2:30 AM

      Or we could be looking at the fact that the Japanese economy has a GDP of $5,964 Billion, and Japanese Imports are $808 Billion dollars, so 87% of the Japanese pay check is spent on domestic goods and services. A big change in import price won't necessarily be reflected in a big change in inflation. Five sixths of the inflation measure is totally unaffected by the strength of the Yen.

      You might expect currency devaluation to be a bigger deal for CPI calculations in a place like Switzerland where imports are closer to half of GDP, and to be even smaller for truly isolated countries like North Korea where imports are less than 10% of GDP.

      Delete
    2. Frank,

      Thanks, but I already thought of that. Over the last two years, the effect of the devaluation of the yen on Japanese CPI, taking into account what fraction of Japanese GDP is spent on imports, is much greater than the actual change in CPI.

      Delete
  9. Why not go directly to the BOJ? Their monthly report comes out in English only a day or two after their Japanese-language version, and includes 40 pages or so of graphs, updated every month. Here's what they say on inflation:

    On the price front, domestic corporate goods prices are rising moderately
    relative to three months earlier, mainly against the backdrop of movements in
    international commodity prices and foreign exchange rates. The year-on-year rate of
    increase in consumer prices (all items less fresh food) is around 1¼ percent.
    Inflation expectations appear to be rising on the whole.

    and then later in the report there is a detailed discussion, with occasional references to the impact of the April 1st tax increase amidst muted consumer indicators since October. But except for year-end bonuses, wages remain flat, and the increase in employment over the last several years is solely among part-time workers. Hence:

    Total cash earnings per employee have generally bottomed out as a whole,
    albeit with fluctuations (Chart 22[2]). Hourly cash earnings of overall employees
    have turned to improve moderately overall as non-scheduled cash earnings and special
    cash earnings have increased (Chart 21[1]). Taking a closer look, as a reflection of
    movements in nonmanufacturing, monthly cash earnings of full-time employees per
    employee have increased marginally, and hourly cash earnings of part-time employees
    have continued their year-on-year increases, albeit at a very mild pace (Chart 21[2]).
    The year-on-year rate of change in scheduled cash earnings, however, has still been
    slightly negative, with the uptrend in the ratio of part-time employees having exerted
    downward pressure and due in part to the effects of the reduced number of hours
    worked of part-time employees (Chart 21[3])

    In sum, less deflation, gradual rises in employment but a looming tax rise of the sort that in 1997 led to a couple quarters of recession led by a collapse in auto sales. But the inflation to date has indeed been energy and imported goods, and once that's removed remains very modest.

    よろしく!

    ReplyDelete
  10. Anonymous11:04 PM

    In fact, the 1997 tax increase did not lead to a technical recession. It was caused by the collapse of a few financial institutions that followed the Asian financial crisis earlier in the year.
    よろしく!

    ReplyDelete