Tuesday, February 16, 2021

Inflation issues


In analyzing whether inflation is coming, Mickey Levy at Berenberg Capital passes along the above graph. These are price indices, so the upward or downward slope measures inflation. Is there inflation? That depends on whether you ask durable goods or services. 

Why are we experiencing durable good deflation, and will it last? Part of the answer is quality adjustment: 

The Bureau of Economic Analysis' (BEA's) official inflation indexes are based surveys of product prices that are adjusted for estimates of quality improvements in new products and services. 

Quality adjustments have had a pronounced impact on the PCE price index of durable goods, which has declined 38% over the last 26 years, despite sizable increasers in many pcorudct prices... (For example, the average sticker price of an automobile has roughly tripled since the mid-1990s, but, because of estimated quality adjustments, the BERA's PCE price index of automobiles and parts has risen only 9% since January 1995.) 

That raises an eyebrow. Yes, a 2021 computer is more than three times better than a 1995 computer, so if you buy it at about the same price, which you do, we can call that deflation. Is a 2021 car really three times better than a 1995 car? Hmm. However, most economic analysis suggests that the BEA understates the quality improvement of new goods, so there is actually more deflation than it seems here. 

Mickey notes 

advancements in the distribution of goods has reduced prices of goods and services. 

So part of the story is Wal-Mart and Amazon. 

Mickey attributes the decline in durable goods prices to productivity improvements.  But if you run down to Home Depot and look around the selection of power tools (guilty habit), however, a glaring contributor has to be China, as well as innovation. Electric motors got cheap, yes, but you can only buy a pressure washer for $50 because of China. That source of price decline may not last. 


... Quality adjustments in services is much more difficult to measure (i.e. medical services) and in meany services, quality has not improved much over time ..the PCE price index for services has risen at an average annualized pace of 2.6% since the mid-1990s, and 2.1% during ...2009-2019

I push back a little. If it is difficult to measure, how do we know quality has not improved much over time? But it is true that medical services, paid and provided by the most arcane system devised by human mind, do not bear the marks of the sort of innovation and efficiency gains we see in durable goods. The quality gain is mostly scientific advance.   

So, looking at inflation, roughly 70% of the price index comes from services. Of that about half is housing services -- yes, the BEA counts housing as a service, not a durable good. Here the BEA tries to figure out the rental value of houses and quality adjust -- and health care where who knows what the prices mean. Many prices are either determined by Medicare or Medicaid, or insurance negotiation which specify those prices plus a percentage. Is an hour of a lawyer's time better than it was in 2010? An hour of university instruction? (No. Worse, I would guess in the humanities, better in computer science.) 

I hope I am unsettling you about just how much we really understand about inflation. That the Fed is obsessing over 1.7% vs. 2% seems a bit strange. Rather than look at a few tenths of a percent variation over time, if we look at variation across goods categories, across methods of accounting for quality change,  across space -- the price level in Palo Alto seems about 3 times higher than it is in, say, Reno -- and across people -- the stuff that people with lower incomes buy has gotten a lot cheaper, and the stuff that wealthy people buy has gotten more expensive, a fact our inequality worriers might note some day --  we might get a truer picture of just how fuzzy our measurements are. 

I don't know what happens to inflation in the short run. But if this graph is true -- services are inflating 2.5%, really are not having much quality improvement, durable goods are deflating 1% due to continued quality improvement -- the long run means more inflation. As durable goods prices decline, unless people buy proportionally more of them, durable goods become a smaller and smaller part of the shopping basket. Then the CPI weights them less relative to services. We have already seen many goods become free such as maps. They then disappear from the CPI. Inflation then becomes the inflation of services only. 


The question of just what inflation is and how it is measured bears directly on the Fed's new plans. The Fed wishes to let inflation grow to something like 2.5%, in order to drive down unemployment, then start raising rates to contain inflation. If you think of inflation as something like the price of gas, that might make some sense. But if inflation is whatever nonsense you get in medical bills, the BEAs imputed rental value of a house, and so forth, that act of delicate fine-tuning seems a bit more fraught. 

Update 2: Chriatin Broda and John Romalis have a nice paper on the last point, that the goods bought by people experiencing low incomes have become cheaper, offsetting in their estimate half of the rise in measured inequality. 


  1. Right---measuring inflation is an art and subjective.

    Automobiles are three times as expensive, but are better made, than in 1995----but then also many consumers can buy used cars, which seem to last for decades. The average age of a car on the road in the US is 12 years.

    Housing rents have exploded along the West Coast and some other parts of the US, a monthly economic immiseration.

    I have to say, the conventional macroeconomics profession is much too concerned with inflation as measured. Perhaps The Reserve Bank of Australia has a better idea, in targeting inflation in a 2% to 3% band, as measured.

    Some central banks target 3%, plus or minus 1%, and the Reserve Bank of India targets 4% plus or minus 2%.

    The Federal Reserve until very recently targeted about 5% unemployment, and a 2% inflation rate. The Federal Reserve branches produced papers that unemployment rates below 5% would trigger higher rates of inflation.

    But in today's globalized economy, perhaps we would be better off if the Federal Reserve targeted 2% unemployment, and 5% or less in inflation.

    A bit of hyperbole on my part to make a point, but worth pondering.

    1. Ben,

      "But in today's globalized economy, perhaps we would be better off if the Federal Reserve targeted 2% unemployment, and 5% or less in inflation."

      Why don't you think we can get 2% unemployment and 0% or less inflation?

    2. Man, I don't know. Maybe the Phillips Curve would become un-prone.

      Anyway, some inflation is no Big Deal. The important goals are low unemployment, even "labor shortages." Persistent labor shortages should do the trick, boost wages.

      Probably higher wages have to be protected by immigration control, and trade restrictions.

      Here is a laugh: "Applying a 25% tariff on all two-way trade would trim U.S. GDP by $190 billion annually by 2025, the (US Chamber of Commerce) group said in a joint study with Rhodium Group, a New York data and analytics firm."

      Read more at: https://www.bloombergquint.com/global-economics/china-decoupling-would-cost-u-s-economy-billions-chamber-says

      Even the globalists say GDP would decline by only $190 billion is the US put on 25% tariffs on all imports?

      The US GDP is $22 trillion or so.

      Even taking the US Chamber estimate on face, that would be less than a 1% reduction in GDP.

      Imagine US businesses and industrialists re-evaluating domestic investment behind a tariff like that. The boost in domestic stability.

      How can anyone invest in US plant and equipment if the ground rules are that subsidized imports are fine?

      Tax wages less, and tax imports and property more.

      Reduce government (including military).

    3. Ben,

      "Man, I don't know. Maybe the Phillips Curve would become un-prone."

      And that would be a bad thing because?

      "Anyway, some inflation is no Big Deal."

      And the same can be said of unemployment - some unemployment is no big deal either. 5% inflation - no big deal, 5% unemployment - no big deal, 10% inflation - no big deal, 10% unemployment - no big deal.

      There are economic policies that get you the best of both worlds - low inflation and low unemployment. As Dale Carnegie would say - "Think win-win".

  2. Why aren't capital market assets considered as subject to inflation? Equities and bonds are at all time highs. Why isn't that driven by inflation too?

    just because the prices of consumption goods aren't increasing, it does not mean that the value of the dollar isn't rotting.

    1. Fat Man,

      "Why aren't capital market assets considered as subject to inflation? Equities and bonds are at all time highs."

      Because there is this misguided notion that lowering the returns on a banker's time (higher bond prices, lower yields) increases the returns on a borrower's time (Real GDP). Yes, stocks are at all time highs, but so are P/E ratios.

      It is frankly the stupidity of US Treasury / Federal Reserve policy (beginning under Robert Rubin / Larry Summers) that combining investment and commercial banking is a great idea. Investment banks justify high stock prices with low bond yields and vice versa.

      What would make more sense is if the U. S. Treasury sold equity where rates of return are commensurate with the output gap - the higher the output gap, the higher the yield, and the lower the price (countercyclical fiscal policy).

      Then you would see money move out of low yielding bonds into slightly more risky government equity. And as a result, lots of good things would happen:

      1. The natural and monetary rate of interest would rise
      2. The trade deficit would fall
      3. Productivity would increase

      The original sin of this mistake in policy begins with Andrew Mellon.

    2. Huh? My comment did not get any issues FRestly discusses. My question is about the evidence of inflation (i.e. what it really is, which the the declining value of the dollar.)

    3. Fat Man,

      The value of capital markets is based upon projections for future output across all firms represented in those markets - they may international and not operate solely within an area where dollars are used.

      The value of the dollar is a measure of current economic output in areas where the dollar is used.

      That doesn't mean the projections for future output (reflected in the capital markets) will be proven correct, all that it means is that people are willing to invest their money to facilitate that future output.

      And that is the point you are missing. There is a missing market that would allow individuals to invest in their own future productive output - government equity.

    4. And that government equity would be dollar centric. If firms want to offshore, fine, they should not be held back from doing so. But if they do, they should expect to find their financing in the places that they relocate to.

  3. "The Fed wishes to let inflation grow to something like 2.5%, in order to drive down unemployment, then start raising rates to contain inflation."

    John, are you forgetting that the central bank owns like $9 trillion in federal debt?
    Before they can even begin to raise interest rates, they would need to find a buyer.

    To this day the Fed is using the Phillips curve trade off between inflation and employment? I thought that crap was discounted a long time ago.

  4. "But if inflation is whatever nonsense you get... the BEAs imputed rental value of a house, and so forth, that act of delicate fine-tuning seems a bit more fraught."

    So if people are marginal buyers of homes that pushes rental costs down. As soon as interest rates rise, that pushes marginal home buyers back into renting pushing rental costs up, forcing the fed (reacting to rising rental costs in the CPI) to push interest rates up even higher.

    Haven't we seen the stupidity of this once before already?

  5. Quality measures seem subjective also. Is my new cell phone higher quality because it's higher data rate or better camera than the old one, or is it lower quality because the battery is not replaceable so it's got a hard lifetime of 2-4 years after which the battery won't hold a charge?

  6. I've come to the conclusion that quality adjustment and hedonic adjustment are nonsense in a world of fast-paced technological growth. In fact, the entire idea of a "consumer basket" falls apart when you consider how some technologies disappear from typical consumption and new ones enter.

    If you took a cell phone today with all of its functionality -- something even children have and take for granted -- through a time machine to 20 years ago, you'd instantly be considered the wealthiest person alive because that cell phone would practically be viewed as advanced alien technology. It's straight-up nonsensical to compare a basket with and without one. Similar examples abound.

    It's such a meaningless construct that is so divorced from its theoretical counterpart that it should be considered an intellectual sleight of hand to use it as any kind of empirical variable.

  7. Great post! To me the most powerful point is that all of this hand wringing over whether inflation is 1.7 or 2.1 percent is utterly ridiculous. Economic is not engineering and aggregate data on inflation is just a very rough and ready wild guess. It's not useless by any means but treating it as though it were some super sophisticated metric that can accurately be measured in tenths of a percentage point is crazy. Another point is that inflation measures - by design - exclude asset prices. The housing number is an imputed rent and not the price of a house. The price of stocks and bonds are also excluded. There are arguments for these exclusions but you also tend to miss parts of the economy where there is lots of inflation.

  8. Update 2: Chriatin Broda and John Romalis have a nice paper on the last point, that the goods bought by people experiencing low incomes have become cheaper, offsetting in their estimate half of the rise in measured inequality.

    Confirming priors I would say. In my opinion, this exacerbates inequality. Presumably a poor person would like to buy many of the same things as a rich person. So the fact that those goods are rising even more than inflation I would argue makes a poor person even poorer.

    The goods and services that a poor person is buying is getting cheaper because the poor are getting poorer and poorer. And vice versa.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.