Comments (9)

• Don Lloyd

• Hedonic pricing is so flawed, both fundamentally and specifically, that its most important characteristic may be to demonstrate that mainstream economists do indeed have a remarkable sense of humor, involuntary though it may be.

  This section is one of my favorites, intermixing cause and effect beyond all recognition :

  "...Todd Reese, the commodity specialist for autos, says he doesn't need hedonics to extrapolate the value of quality changes, because auto makers present him with a list of changes to the car and the corresponding prices. Still, Mr. Reese must make some tough calls as he does his job. For instance, he recently considered a 2005 model in which the sticker price went from $17,890 to $18,490. The manufacturer cited an extra cost of $230 to make antilock brakes standard, while it said it saved $5 by dropping the cassette portion of the CD player...."

  Regards, Don

  
  

• Published: May 19, 2005 2:28 PM

• Bill R.

• The purpose of hedonic adjustment has always been political, even though the justification appeals to (some) math nerds. Any discussion of the “meritsâ€? of hedonic adjustment is therefore moot. Hedonic adjustment serves the same purpose that “owners’ equivalent rentsâ€? and “substitutionâ€? or “chained CPIâ€? does – to lower the reported number. Note that most elements of CPI have a hedonic adjustment, but I once spent several hours searching BLS.gov for a list of the items with/without adjustment and couldn’t find it. It may be there, but they certainly don’t make it easy to find.

  If the government really wants to measure “inflation� – well, they already do measure inflation, with perfect accuracy! You can find that measurement at http://research.stlouisfed.org/fred2/ under “monetary aggregates.� Back before the CPI started getting changed, you could see a very direct link between the monetary aggregates and CPI, but CPI lagged M by about 18-30 months. Now that CPI has been “adjusted� it tracks M less well. But that’s not because the world has changed, only because the government has on a new pair of glasses.

  If the government really wants to measure the cost of living increase in nominal dollars, they should (among other things) avoid hedonic adjustments, for a variety of reasons. In a stable money system the improvements in quality for any given constant price level are expected; this is a function of the marketplace. Businesses compete on quality, price, and service (usually only two at a time). Hedonic adjustments are immaterial to most consumers’ needs. The fixed income senior needs a new car, having worn out the old one over a period of fifteen or twenty years. The new car has “hedonic improvements� like direct cylinder fuel injection, multiple air bags, improved emissions, OBD II computer chips to alert him to emission problems, anti-lock brakes, tire pressure monitors, etc. But Mr. Senior doesn’t notice these improvements … he notices that the cost of new car is much higher than the last one he bought, though. Hedonic adjustments are not “optional.� Assume there was a car company that (in blatant non-compliance with multiple safety and environmental laws) produced a new exact copy of the 1974 Super Beetle in all its air-cooled, carbureted glory. Assume further that the BLS informed me, through CPI numbers, that this “new Beetle� cost as much as a new Camry does. But this doesn’t add up! The new Camry costs twice as much as the “new Beetle� in actual dollars yet the government tells me the cost is the same because the Camry is “improved�??? I know which one I would buy. But I don’t have that choice. They don’t make cars like they used to, or washing machines, or TVs, or radios, etc. And since I don’t have the option to trade down in “quality� for a price break, using a hedonic adjustment to tell me the price is unchanged is like relieving yourself on my head and telling me it’s raining.
  

• Published: May 19, 2005 4:04 PM

• Jay S.

• Good comments here.

  Help me out, if you remove the $ value of improvements out don't you have to put product unimprovements back in? Or am I confused? Unimprovements like car bumpers that can’t be used to bump. Cars the cost $300 to tune up, in the old days I could do it myself with a strobe light socket wrench and feeler gauges. Plastic fenders costing $2500 to repair, compared to some metal filler and hammer?

  I understand the justification is to justify the long held belief that there must be 0.5 to 1.0% improvements in the CPI.
  

• Published: May 20, 2005 8:05 AM

• Bill R.

• CPI is a lie! Since the 1980's the Revisionistas have been changing the definitions for CPI, through hedonic measurements, the substitution of "owners' equivalent rents" for housing prices, among other possible changes. Now the "implicit price deflator," a little-known tool used to turn nominal dollar GDP into "real" GDP, also has chain-weighting and substitution. As a result of which the "inflation" rate measured by it is even more suppressed.

  There is a dearth of information available about the black art of CPI, IPD, etc., and there is also a great deal of brainwashing going on along the halls of academe about the reasons behind the methodology. But, with publicly available facts, I will attempt to convince you that CPI (along with GDP) is indeed a lie.

  From what I’ve been able to gather, the “hedonic measurement� craze began in the late 1970’s with automobiles and has expanded into all sorts of appliances. The substitution of “owners’ equivalent rents� for housing prices was implemented in Jan 1983. Also there have been various tweaks to GDP measurements.

  For the period 1960-1982, the R-Squared for M3 YOY (year over year) change to Nominal GDP is 12.9% and to Real GDP is 0%. You expect the nominal number to track money supply to some degree, since it is a consumption measure – no problems here. But the good news is that Real GDP didn’t track M3. So whatever measures the government used for the IPD to turn nominal into real GDP, it at least was not predictably biased by money supply.

  Turn to 1983-2003. The R-Squared for M3 to Nominal GDP is 6.8%, and to Real GDP it is 18.1% (with R>0, so increased M3 = increased real GDP). Something has changed here, folks. Either (1) the fundamentals of money supply, (2) the measurement of nominal GDP, or (3) the measurement of inflation, or some combination of these, changed.

  I checked the inflation measurement contained in the Implicit Price Deflator. The R-Squared to M3 is 10.7% with positive correlation from 1960-1982, and is 4.3% with negative correlation from 1983-2003. So let me get this straight … using government data, the correlation between increased money supply and higher prices is NEGATIVE today, when it used to be POSITIVE?????

  OK, let’s check the CPI inflation measurement. Given the Austrian theory of the business cycle and what I know about how the Fed “creates� money, I will compare the M3 change of two years ago to the CPI inflation of the current year. This allows time for the money to “trickle down� into the general price level. Before 1983, R-Squared is 38.6% and positively correlated, after 1983 it is 0.3% and negatively correlated. Again, using government data, the correlation between increased money supply and higher prices is NEGATIVE today, when it used to be POSITIVE?????

  OK, let’s recap. Two decades of evidence that increasing money supply drives up prices and does nothing for Real GDP. Followed by two decades of evidence that increasing money supply drives DOWN prices and increases Real GDP. In the interim, we know that certain “adjustments� have been made to the IPD, such as wider use of hedonics, substitution of rent rates for housing prices, and “chained� substitution methods.

  I don’t know what you believe, but I believe:

  1. That the rate of increase in dollar prices for 2005 is totally contained in the monetary aggregate increase for 2003;

  2. That the government is lying to us for political gain;

  3. That CPI is part of that lie.

  Go to BLS.gov and download the CPI changes, year over year. Graph them. From 1960-71 it topped 6% ONCE. From 1971-1981 it topped 6% EIGHT times. From 1982-2003 it topped 6% ONCE. However, we've four years of double-digit M3 growth this recent two decades, as well as the lowest "real" interest rates over the entire half-centure. Does it make sense to you that inflation finally got whipped? Or does it make more sense that inflation is being papered over?

• Published: May 20, 2005 2:16 PM

• Daniel

• I'm studying economics in general (especially austrian) and I don't understand one thing about the money supply and the comments above: should at least some of the money supply growth be explained due to natural population growth? I mean, if the number of people increased and the supply of money did NOT grow, wouldn't we see the dollar increasing in real value, since more people would be chasing the same amount of dollars? Thank you!

• Published: June 3, 2005 8:47 AM

• Gil Guillory

• Daniel,

  Yes, other things being equal, the purchasing power of money would rise with an increase in population. But what is not clear is whether an increase in the money supply is somehow "necessary" or conduces to social welfare.

• Published: June 3, 2005 9:09 AM

• Michael A. Clem

• Well, let's see: if the money supply is increase, causing inflation and a decrease in the purchasing power of money, then those that have saved money have lost some of the value of their product.
  
  What would the reverse situation be? If the purchasing power of money increases due to population growth, wouldn't those who held assets instead of money be losing some of the value of their product? And those who held money would be getting more than the value of their product (when they spent the money)?

• Published: June 3, 2005 10:58 AM

• Daniel

• I see - so what I get from the 2 comments ago is that saved money is an investment like any other, if the government prints more money, it would be artificially reducing the value of my savings - while I should be getting the benefit of a population growth since I saved it - like holding a stock for an oil company before oil consumption goes up for whatever reason.

  So the best is having the government out of money decisions and let to people judge their savings rates by their time preferences and expectation of future value of their product (money), right?

• Published: June 3, 2005 2:51 PM

• Bill R.

• Imagine a fixed money supply and fixed population. As the economy progressed, the only way businesses could compete for increased profit dollars would be to improve efficiencies. This improved efficiency would tend to LOWER prices.

  During the late 1800's in the U.S., with the dollar fixed to gold, monetary supply inflation had a natural limit, i.e., the amount of gold stored in reserve by the government, so it increased a tad over 2% per annum. Population increased at a higher rate. In that period, the data suggests that the actual price level was quite stable.

  Don't let any "economists" fool you into thinking that saving or hoarding cash hurts the "economy." Hoarding actually depresses the general price level. Every dollar you hoard is one dollar less competing for resources I may want ... now. Saved money is an investment. In an inflationary environment (monetary inflation that is) it is a losing investment.

  The devaluation of a currency is a tax on holders of that currency. This works both for U.S. citizens and for foreign countries.

  Since the US$ is ipso facto the reserve currency of the world, in large part thanks to OPEC agreements in the '70's to exchange oil only for US$, foreign governments have large US$ reserves. When the US devalues the dollar, it is a transfer of wealth from these foreign governments, since more foreign assets are denominated in US$ than in any other currency - they NEED dollars to BUY OIL from OPEC.

  What is a tax? It's when government decides to take your money and spend it the way they want to. Inflation of the dollar supply does this. "Easy Al" and his "green meanies" don't sneak into our houses every year on Christmas to give us an extra $1.50 for every $20.00 in our wallets and checkbooks. No, if they did that, the general price level would rise in all places fairly equally and no one would get any benefit from first use of the money. But imagine a counterfeiter that made dollars so perfect that no one could tell the difference! That person spends the extra dollars first, and can outbid all other usages for capital based on the counterfeit funds. He gets "first use." By the time those extra dollars "trickle down" into the general wage pool, Joe Six-Pack has been paying higher prices for 2-3 years. This is what the Fed does. The "first users" of the government's newly printed (or, digitized) dollars are the banks and institutions, the military-industrial complex, and the government itself, of course. Anyone first in line at the feeding trough gets the pick of the pork, and everyone else gets "sloppy seconds" or a tax on the currency they hold.

• Published: June 3, 2005 10:57 PM