They have to calculate the inflation rate on products like this:
The iPod will be six years old next month. The newly released iPod Classic with 160 GB of memory is $50 cheaper, holds 40 TIMES more songs, plays color videos and displays photos than the original. It is smaller, lighter and has a better battery. I wonder how the BLS takes account of the quality differences when measuring the price index and inflation.
Hi -
Doesn't the word hedonic deflator mean anything?
This is basic stuff if you have been paying attention to the reasons behind the move to chain-weighted indices and the kind of work that the Fed does on the G17 report on industrial production: hedonic deflators are used to take into account changes in quality.
In a slow and fairly static economy you don't need to take into account these kinds of things: in a dynamic and fast-growing economy, you do. Back when I forecast the US computer industry, either you went with the hedonic deflator or the US computer industry, with its large double-digit nominal growth rates, ended up representing around 110% of GDP before the price basket could be updated. Things change way too fast for old-school statistical analysis.
For that reason, classic, stick-in-the-mud economists hate chain-weighting and hedonic deflators, because the meaning of a real price ceases to make sense to them any more: a hedonically deflated chain-weighted price index cannot be used the same way that a constant basket deflated price index has been used in the past. They're two complete different things.
And the problem isn't just consumer electronics: the automobile industry also shows massive hedonic changed. Then again, perhaps in the next round of industrial classification schedules revamping, they might want to put automobiled into the consumer electronics sector anyway... :-)
That is an interesting problem.
You get more from today's ipod than the old ones. But isn't the same thing true of a kilowatt of electricity? Think of all of the things you can do with it today that you could not do with a kilowatt of electricity years ago.
If all that happened between year 1 and year 11 was an increase in the use of air conditioners, electric heaters and refrigerators, you could say that the kilowatt of electricity is the same product, just more in demand, thus it became more costly. In that case, it contributes to inflation. But if in the interim it also began powering TVs, computers and iPods, you can argue that it is an intrinsically more valuable item, and an increased cost of a kilowatt might be offset by its increased value.
I think such an argument would be bogus. I also think that the same aspect of that bogusness applies to the ipod, but I'd probably need to take an more than econ 101 to be able to explain how.
Redefining inflation is a fun political game that anyone can play. Let's say that oil prices are rising, but you want to say inflation is tame. You just define "core inflation" to exclude oil. Hey, Presto! The President's economic team is doing a helluva job.
When they put in "hedonic deflators" for stuff getting better, do they also mput in "hedonic inflators" for stuff getting worse? You know, all the things where service standards decline, or where the burden is pushed over onto the consumer.
Anyway, is more tunes good when the music is mostly dire?
Njorl wrote: If all that happened between year 1 and year 11 was an increase in the use of air conditioners, electric heaters and refrigerators, you could say that the kilowatt of electricity is the same product, just more in demand, thus it became more costly.
...and even in those examples it gets even more complex. Compressors have become more efficient over the years, and a refrigeration system installed now will often move more heat per unit of consumed energy as compared to one installed in the mid-1990s (and definitely much more than one installed in the mid-1980s).
Njorl - The 1 KW hour of electricity can be put to more uses today, but its still pretty much the same thing. Its more useful because of changes in other items. To the extent you hedonicly deflate the inflation adjustment it would be do to changes in those other things not the electricity. If you also adjust for the increased usefulness of the electricity your double counting the change. In the case of the IPOD you actually get a different and better IPOD today for the same or less money, the difference is in the IPOD itself. If buying music from Itunes gets cheaper, that might make the IPOD more useful (at least to people who use Itunes rather then ripping the music, or getting it from PtP) but the deflation would be in the price of music not a deflator for the Itunes.
It's possible to get a handle on this.
http://www.klenow.com/BKQ.pdf
The intuition is that wealthier people buy both more of any particular good than poor people, and better instances. The ratio of the extent to which they choose to buy more vrs. better goods can be measured by comparing rich and poor people at one particular time.
Then, in a future time when efficiency improvements have made everyone's [say] iPod dollar go further, we can assume that people are choosing to take advantage by buying both more iPods and better ones in about the same ratio. Therefore, when we see how many more they buy we can tell how much better each one is.
iPods is both a good and bad example. Bad, because it's been available for only a short time [although it compares to CD players or portable casette players]. Good, because it calibrates the technique over a period of time which is short enough that not too much inflation has taken place and certainly there isn't that bad a disagreement as to how much [except that the paper is a 2001 paper and does not include the evolution of the iPod over time].
Will the BLS do anything about this? No. It's a hard explain, and therefore demagogable by the opposition party to the administration that tries.
-dk
Megan_McArdle: To answer your question, the way I'd do it, and the way I hope they do it, is to figure out how much you'd have to spend to get the functionality of the current iPod, six years ago, and compare that to the current price.
You would probably have to buy:
1) original iPod
2) 156 GB external storage + input to iPod
3) Special case to approximate current compactness and convenience.
4) Color screen converter.
That would probably run you ~$3000 back then, so the change from $3000 to the current price is its effective change in price.
What I've described can be generalized to compare living conditions across times and places. You simply compare how much unskilled labor it costs to obtain the *functionality* of some good between the two times/places.
dearieme is right, though: If you count quality improvements against inflation, you have to count quality debasements toward inflation. And I think the latter are pretty much ignored in their methodology...
Here is a paper describing some aspects of hedonic adjustments.
Such adjustments must be made if one is going to get a true picture about the cost of living. Products come and they go, and in this process they almost always change in quality- usually for the better, and when they change for the worse, it is almost always because the item is losing its customers and becoming extinct. Computers are everyone's favorite example of an item that has changed in quality by several magnitudes to the better, while actually falling in price in nominal dollars. However, one should look up the cost of automobiles from the 70s in the dollars of those days, adjust for CPI to price it in today's dollars, and ask themselves whether autos are really more expensive today than they were then. A Ford Pinto from 1976 would cost almost $11,000 in 2007, but you can buy a number of brand new subcompacts today for that amount, but have a car far superior to the Pinto (I know, because my first automobile was a 1976 Ford Pinto).
And just in case consumer electronics or cars aren't enough of a discussion tarpit, just consider how much better healthcare is (in terms of drugs and treatments available) than it was 10 or 20 years ago...
Mike_Earl: Really? Have doctors started making house calls again? Are free clinics as good as they used to be? Is it less arduous to get painkillers?
"Have doctors started making house calls again?"
I imagine you could get some doctor to go to your house if you had enough money. I can see why "core inflation" is used. Things that get essentially priced out of the marketplace - like house calls, new inventions, obselete items...they really complicate things.
Maybe you should follow the advice I gave Russ Roberts -- call the BLS and ask them how they do.
they will be more then happy to explain it to you and you might find out that they actually do a very good job of handling things exactly like this. This particular example would not be difficult--compared for example to a computer or an auto.
Moreover, you might find out that the iPOD has such a small weight in total consumer purchases that even if they completely ignored it it would not have a significant impact on the reported inflation rate.
It is easy to make armchair arguments and fire off blog posts (or "analysis" for your clients) with grandiose claims and lame statistical analysis in order to criticize official statistics (which, of course, must be produced under a given budget and in a timely manner).
If your ideas are so brilliant and only morons at statistical agencies can't see them, write them up properly and send them to an academic journal for review.
I think you may be surprised at the level of sophistication in the literature, and that your ideas are not workable. But if not, you might actually do some good for the public ... rather than undermining confidence in official statistics.
Njorl ... if something gets priced out of the marketplace ... THAT'S A SIGN OF INFLATION!
Second, my methodology easily handles new inventions and obselete things by focusing on unskilled labor cost of achieving that *functionality*. We may not use punch card computers today, but we can figure the price to obtain that level of *computational output*.
Third, core inflation doesn't handle any of the issues you mentioned; it merely subtracts out food and energy on grounds of their volatility.
"Njorl ... if something gets priced out of the marketplace ... THAT'S A SIGN OF INFLATION!"
That's an easy thing to just say, but if nobody is buying it, what does it cost? I think inflation calculations should consider volume to be most accurate. The prices of buggy whips and house calls don't affect inflation because nobody buys them. The reason nobody buys them are exactly opposite - one is worthless to the prospective buyer, the other is exhorbitantly priced to the prospective buyer, so neither is bought.
Person:
Yes, labor gets more expensive as we get richer, and medicine is over-regulated. (And a late-80s PC has a better keyboard than a modern one...)
That said, a randomly selected American asthmatic or diabetic or high-blood-pressure patient is almost certainly getting more effective care than he would have gotten 20 or 30 years ago.
Mike_Earl: randomly selected American asthmatic or diabetic or high-blood-pressure patient is almost certainly getting more effective care than he would have gotten 20 or 30 years ago.
Sure, just not someone who wants the doc to pay a quick visit or who has stabbing chronic pain he wants to get rid of. I'm in the latter category, and I know how hard it is. Don't lie to me about how much better my medical care is.
Njorl: First of all, that wouldn't apply in the methodology I advocated. If a horse/buggy were in the CPI basket and they became obsolete, you could replace it with someone in use that is functionally equivalent, such as a golf cart.
Second, it's not problematic to de-weight things as they fall out of use, although it's again less of a problem if you look at functional equivalents.
Third, when you said "priced out of them market", that expression implies that people stop buying it because it becomes expensive, not that it becomes expensive because people stop buying it [and returns to scale disappear].
Fourth, doctor house calls' expense has little to do with how many of them do it. Any doctor could start it without the consent of other doctors, and would not be burdened because no one joined him.
Including inflation into the equation makes Moor's Law even more impressive?