Own Comment Rescue, Inflation Edition

Continuing with the recent, curious trend of deeming my comments worthy of being highlighted in posts of their own, our friend Murali sent me an email urging that I extract a comment I made (in another post that had extracted a portions of comments of mine and Brandon Berg’s, no less) about inflation and technological innovation and make it an Off The Cuff item. So I’m doing that. I’m also including a response from J@m3ez Aitch that I thought answered my question quite well (my comment is, after all, a request for explanation). So here those are.

First me (in my slightly arch smarty-pants tone where I try to Talk Pretty about things I really have not the first clue about):

A question for the economists (credentialed publicly or in their own minds) out there:
Relating to James’ & Rod’s discussion above of the competing “tropes” of increased, or at least not as much decreased as claimed, buying power of the non-wealthy classes and the related idea that much of that increased buying power relates to a limited set of rapidly-evolving products arguably mostly clustered around non-essential or luxury facets of life (entertainment, portable hi-fi real-time person communication or computing, etc. [‘essential’ is certainly in the eye of the beholder here, but I digress…]) — a question about inflation:

Being that it has been said that much of this increased buying power is related to items a person can’t eat or live in or drive or that eventually produces a professional credential, is it reasonable also to wonder whether it has been (sufficiently) accounted for in official measures of inflation? I.e., if the U.S. CPI “basket of goods” doesn’t sufficiently keep up with how much more computing and connecting power can be bought for $100 or $200 for the hardware and $100/mo for the data (or what have you), then is it possible that in some sense (though perhaps not in an important sense from a monetary policy perspective, as there continuity of comparison, not representation of the world as it’s being experienced, may be the primary imperative), common measures of inflation may be considerably overestimating a “true” economy-wide inflation level in which prices of comparable goods are compared over time?

Or do I just have a category error here, where I am thinking that inflation as a concept is meant to say something about actual purchasing power across an economy over time, when in fact, it is just meant to give information about how the money supply (or something) is interacting with (for calculation purposes, a certain “basket” of) goods in the economy to create price trends, and that the ability of units of currency to command goods other than those representing “goods in the economy” (i.e. goods not in “the basket”) for purposes of the calculation is just pretty much unrelated to the calculation (and the concept of the value being calculated), and so this obviously won’t reflect all the (changes in) facts about a currency’s (or a person’s or class’s) purchasing power over time?

I would appreciate if I could have this moment of conceptual confusion cut short.

And then J@mez, helpfully concrete as ever:

I don’t think you have a category error. I think you’re just confused because you see two things going on in our inflation calculations that you can’t easily reconcile, and that’s because in fact they’re not easily reconciled.

Our inflation calculations do say something about purchasing power, but not everything about it. If the cost of items we buy goes up, then our inflation calculation reflects an actual decline in our purchasing power. If the costs goes down, the inflation number reflects an actual increase in our purchasing power.

But many economists think we overestimate inflation in two ways. One is as you suggest, improvements in quality. If I can buy better quality for the same amount, my purchasing power has actually gone up; or if that gadget got cheaper as it got better, then my purchasing power has gone up more than is reflected strictly by the numbers.

Another way is through the shift to substitute goods. On an industrial scale, this can be a shift to one material to another, as the beverage can industry shifted from tin to aluminum. On the consumer scale this is often discussed in terms of shifting from beef to chicken when the price of beef goes up. If beef prices increase, so people shift to buying chicken, they still get plenty of protein, so their purchasing power hasn’t declined as much as just the change of beef prices might suggest.

But we can object to these. If you’re ambivalent between beef and chicken, then substituting one for the other is fine, but what if you like beef, and don’t really like chicken? Then doesn’t a price-induced move to what for you is a lesser-quality (in taste, if not in nutrition) food actually reflect a decline in purchasing power? And what if you don’t want all that increased technology, but can’t buy anything less? For my part, I’d like to be able to buy a newer car with a manual transmission and hand-crank windows–I just pulled a door panel on one car to put a piece of wood in there to hold up an electric window that stopped working, slides down on its own, and isn’t worth the cost of fixing, and am now being hit with an $1800 bill to fix the automatic transmission on another car, and these are items I don’t even want in a car! Or think of Alan Scott’s $300 cell phone comment.

These critiques tend to be “liberal” critiques. You won’t hear them as much from conservatives. And I think if we stick seriously to the concept of subjective value, then as a matter of intellectual consistency we can’t simply dismiss them. The problem is that we can’t measure subjective value very directly, whereas we can measure price directly, and for “objective,” or perhaps I should say “technological” quality, we can at least estimate it better than individual quality.

And for making official determinations of inflation, we’re kind of forced to use the more measurable things. Which is fine as long as we remember we’re not including in our numbers that which we can’t measure…but once we remember that, what can we really say with any certainty about that which we can’t measure?

Additionally, the choice of what to include in the basket of goods is not objectively determinable, so it’s always a bit political. If there’s actually a large scale shift from beef to chicken, should beef remain in the basket or not? Energy prices seem obviously an important thing to consider, but in the short run there’s enough volatility that it’s not clear they actually reflect real changes over time.

In a nutshell, probably the only thing about inflation measures that the economics profession is in full agreement on is that despite its value it’s inevitably problematic. And I think you’ve essentially recognized where the problem lies.

In retrospect, I think this is more of an Aitch comment rescue done by Drew, encouraged by Murali, than it is a Drew own-comment-rescue. That was some high-level and on-point ‘splainin’ from a great teacher right there. There are also further helpful comments in the thread, including from the aforementioned Brandon Berg. Read the whole, as they say. Many thanks all around!

Update: Since this particular question has been pretty well dealt with by J@mez IMO, perhaps we should make this an open thread on any of the questions raised in the series of posts and comment threads that can be followed back via links from this one to Vikram’s “ Share