16 Feb 2011

PPI Continues to Alarm, So BLS Will Change It

Conspiracy 19 Comments

Bob Wenzel does a good job explaining the alarming trend in the Producer Price Index. I strolled over to the BLS website to see for myself, and was nonplussed by the front-page explanation that they are introducing a new “experimental aggregation system” for the PPI:

Currently, the Bureau of Labor Statistics (BLS) uses the stage-of-processing (SOP) system as the key structure for analyzing producer prices. This system aggregates commodity price indexes for processed and unprocessed goods and is organized into three stages: finished goods, intermediate goods, and crude materials for further processing. Over the past 20 years, the BLS has expanded Producer Price Index (PPI) coverage to include price indexes for many service and construction activities. To expand the scope of coverage for the PPI, BLS recently developed an experimental aggregation system that includes price changes for goods, services, and construction sold to all portions of final demand and intermediate demand, based on information from the Bureau of Economic Analysis’ Input/Output accounts.

Huh that’s kind of interesting. I wonder if the new numbers will make reported PPI inflation higher or lower than the old approach?

UPDATE: I should point out that right now, it apparently truly is “experimental”:

“The aggregation system described here is in its experimental phase, and the PPI program is seeking feedback from its users.”

19 Responses to “PPI Continues to Alarm, So BLS Will Change It”

  1. Captain_Freedom says:

    Don’t worry, when computer servers become too expensive on account of inflation, and more and more businessmen resort to investing in used abacuses, the PPI basket will change to reflect that, and inflation will be contained.

    • bobmurphy says:

      And have you tried out the new abacuses? They are 50% more ergonomic than the ones Archimedes used. It’s the Deflation Millenium.

      • Silas Barta says:

        And don’t forget the Hindu-Arabic place-value notation system our Brother in Italy introduced. It’s a big improvement over Roman Numerals, allowing us to get more use out of our abacuses, and thus decreasing the effective price of computation.

        One you account for this kind of hedonic adjustment, you can see that the gold our explorers are bringing in from the New World isn’t really causing inflation.

        • bobmurphy says:

          I think we all have to agree that the cleverest comments occur on Free Advice.

          The most deranged, out of touch, and narcissistic comments too–but the cleverest, let’s not forget.

  2. Daniel Kuehn says:

    It would be interesting to dig into the discussion of this and see when it started. I’m guessing these things don’t just start from scratch – it’s probably been in the works and under discussion for a while. The problems posed by price indices aren’t exactly a state secret.

    • bobmurphy says:

      I am glad you are always here to play Devil’s Advocate Daniel. Keeps me honest.

      If they started this experimental program in September 2008 I am going to call Fox News though.

      • Daniel Kuehn says:

        Oh, I wasn’t even trying to play devil’s advocate in this case! (and I should hope I don’t have to be one all the time).

        Your update suggested to me that you didn’t think this was a genuine conspiracy to hide anything either, and I’m just noting that the price indices are among the most suspect of fedeal statistics. Not because of foul play, but just because it’s such a hard thing to track in a reasonable way.

        • Dan says:

          Are you aware of any adjustments to the indices that caused their inflation readings to be higher than they previously were?

          • Desolation Jones says:

            Core CPI was showing about 1.5% inflation in 2009 while the regular cpi with food and energy was showing literal deflation.

          • Dan says:

            You’re comparing apples to oranges. Show me where an adjustment to core CPI caused core CPI to rise higher than it would’ve without the adjustment or show me where an adjustment to regular CPI caused regular CPI to rise higher than it would’ve without the adjustment.

          • bobmurphy says:

            Desolation Jones, I know you probably won’t believe this, but I think the financial press focused more on the headline numbers during that period, shouting “deflation! aghhh!” and then they switched over to sterilizing the impact of the headline numbers by pointing out the moderate core numbers, once the headline started rising again.

          • Desolation Jones says:

            Murphy, unless there’s a another bigger conspiracy of the BLS conspiring with the media, the media talking about the most sensationalist number isn’t really surprising. After all core inflation is at record lows. On the other hand, I’m sure the media didn’t say a word about the low core inflation when the headline inflation rate was really high in 2008.

            Dan, I thought using core inflation as an example was within the spirit of the conversation since core inflation tends to be one of those things people like to say was invented to underestimate inflation.

            If you want something directly with headline CPI, December’s seasonally adjusted CPI (SA is always the headline) showed an unusually high .5% inflation while the unadjusted showed .2%.

          • Dan says:

            “Dan, I thought using core inflation as an example was within the spirit of the conversation since core inflation tends to be one of those things people like to say was invented to underestimate inflation.

            If you want something directly with headline CPI, December’s seasonally adjusted CPI (SA is always the headline) showed an unusually high .5% inflation while the unadjusted showed .2%.”

            Other than the period you mentioned has core inflation consistently shown a lower number or higher number than regular CPI?

            I’m not sure you getting what I’m asking since you bring up SA numbers. Have they made changes to CPI, core CPI, PPI, core PPI, etc. that caused the readings to be higher than they would have been under the old system. For example, if we were calculating CPI the way they did in the 70’s or 80’s or 90’s would it be showing a lower or higher number?

          • Desolation Jones says:

            Again, seasonally adjusting is the spirit of the government fudging numbers to underestimate inflation. Okay, you mean changes to the way cpi calculated compared to an older way?

            Houses used to be part of CPI, but it was changed to homeowner’s rent equivalent. Inflation would have been lower the past five years if it wasn’t for this change. You think I’m just cherry picking this period because of the bubble pop? Lets look at how it affected CPI when the home the change was introduced.

            http://i.imgur.com/GvPJT.jpg

            “It is often incorrectly assumed that the introduction of OER lowered the growth rate of the shelter index in the CPI-U. Chart compares the CPI-U with the CPI-W, which continued to employ the old homeownership approach until January 1985. Primarily because interest rates moved sharply downward during 1983 and 1984, the increase in the cost of homeownership as measured by rental equivalence in the CPI-U was greater than the increase as measured by the old approach used in the CPI-W.”

            http://www.bls.gov/opub/mlr/2008/08/art1full.pdf

          • Desolation Jones says:

            Meant to say:

            “Lets look at how it affected CPI when the change was introduced.”

          • Dan says:

            How is that your response to, if we were calculating CPI the way they did in the 70′s or 80′s or 90′s would it be showing a lower or higher number?

            I level a criticism against how the BLS calculates inflation and you post a link with the BLS saying no we don’t? Are you claiming that if calculated the CPI using 1983’s measures that inflation would be lower today? If not, what’s your point?

            http://www.shadowstats.com/alternate_data/inflation-charts

          • Desolation Jones says:

            You asked:

            “Are you aware of any adjustments to the indices that caused their inflation readings to be higher than they previously were?”

            I linked the graph that showed that the 1983 housing adjustment made cpi higher than what it previously was. How does that not answer your question?

          • Dan says:

            Hey, I’ll concede that these guys might make a gaffe that doesn’t make the number go lower like they would like it to sometimes. I would only concede that it was a gaffe though.

            The spirit of the question was do they manipulate the number higher or lower? You have to go back to 1983 for an example of where it temporarily reported higher inflation than otherwise from the previous measures. So unless you want to show me measures being used in 1983 that would report lower inflation than today’s measures, I don’t think this is a point for you.

        • bobmurphy says:

          You’re giving me too much credit. I still think it’s a conspiracy (in the way you are using that term here), but I wanted to clarify that the PPI hasn’t yet been altered, at least according to their statement. I.e. my post title and discussion would have led the reader to believe the BLS already implemented the new technique in the numbers reported to the press for “PPI.”