New York Diary

The Cost of Things

Tim Harford wrote an interesting piece on inflation for Friday's Financial Times (which, oddly, you can't read on their site unless you have a subscription, but which you can read on slate for free). The basic claim (you should read it yourself) is that it's really hard to measure inflation, if not impossible. This is for two reasons: the same object in times past and today is not always equally desirable, and the same object 100 years ago and today is generally not actually the same when you get all nitpicky about it. So, even if you can say that a hat today costs 50% more in real dollars than it did 100 years ago (and this itself is hard enough to establish), this doesn't take into account the fact that a hat was far more important socially and culturally at the turn of the 20th century. It's true that a hat may be more expensive now, but we just don't care, because owning one is hardly a necessity for professional life. The cost of living is not more expensive today because hats are more expensive. The flip side of this is that objects themselves change and improve. We may be spending 10 times as much on the average warship today relative to 1776 dollars, but the average warship can do far more than 10 times as much damage. Cell phones may be more expensive in real dollars than the old rotary phones ever were, but cell phones can go anywhere. (Heck, normal, off-the-shelf, land-line phones may be more expensive, too--but they're also probably wireless, equipped for conference calling and multiple lines, come with hold buttons and caller ID, and have a volume control. Perhaps you're not paying more because the value of the dollar went down, but rather because the quality of the phone went up)

All of this makes sense to me, and complicates the idea of measuring the changing value of the dollar. I think Harford opens the door to a further aspect of the inflation problem, though, without ever really considering it well.

We measure inflation by comparing the cost of a static basket of goods over the years. If the entire basket costs $2,000 in 2000 and $3000 in 2005, we've just seen a 50% inflation rate over the last 5 years. The basket is big; it contains everything from college tuition to men's shirts, uncooked long grain white rice to haircuts to instant coffee to sports equipment (and a million other things, which you can see here). But just as not all of these things were equally important to daily life 5, 10, or 100 years ago, not all of these things are equally important to people of different social classes today, nor do they require an equal proportion of different people's incomes. Over the last century, the cost of technology has generally been decreasing. The cost of food and rent, however, has generally been increasing. If you're the kind of well-to-do person who owns (or even who has inherited) your home and who makes enough money to have to spend only a small percentage on food (freeing up a large amount for, say, flat-screen TVs and new computers and other high-tech toys), the inflation rate on the things that you buy has been very low. If, however, you are the kind of working-class person who spends the bulk of your income on rent and supper, then you're feeling inflation very much indeed. If you are a businessman, the increasing real price of suits is troublesome. If you're in construction, the decreasing real price of boots is encouraging. For the former, wardrobes are becoming more expensive every year; for the latter, they are becoming cheaper.

To try to account for all of this, the basket of goods that is used to determine inflation is meant to reflect some average of what Americans purchase (in rural and urban areas--there are two different baskets that divide along these lines). But few Americans actually purchase the "average" number of computers, for example. Rather, Bill Gates owns an awful lot of 'em (as do even the normal upper-class families--a laptop for dad and another for big sister, a PC for mom and little Jimmy to share, plus sis's old computer in the back room and that new laptop on order now that Jimmy is old enough to start typing his assignments for school), while people on the other end of the economic spectrum own one computer or none at all. Similarly, it's not the case that most Americans include one package of Ramen noodles--or 1 ounce of fresh cheese--in their shopping baskets each week. Rather, some people are eating a bunch of Ramen, and others are eating gourmet cheese. These people may indeed overlap--the same person might just LOVE both chicken-flavored noodles and fresh Parmesan--but it seems unlikely that anybody is regularly buying a single package of Ramen for 20 cents, or cheese in anything less than 5- or 6-ounce chunks. In other words, the inflation rate is being calculated on a basket of goods that reflects not the average American, but rather the nonexistent American.

Maybe we would do better to try to calculate inflation based not on the mean (= number of things / number of people, so that 100 computers / 50 people means each person has to have bought 2 computers even if in actuality one person has 35 computers, 60 others have 1, and the remaining 5 have none at all), but on the mode (= the greatest number of people in today's population have 1 computer, therefore we include only 1 computer in our basket of goods, even if that one guy had to have bought 35 of 'em for this to work out). Better yet, maybe we ought to look at different consumer price indexes for members of different tax brackets, much the way we already recognize that urban consumers have different needs and desires from their rural counterparts. After all, just as a Manhattanite can be presumed not to need a tractor and large quantities of fertilizer, and not to own vast swathes of land, while a rural Iowan can be assumed to have a lesser need for the Club (for the car) or a gym membership, America's richest families feel the effects of inflation when it hits investment returns, new car prices, technology, jewelry, wine, and other luxury items, while our poorest families most strongly feel the decreasing real value of the dollar when they discover that a greenback will buy less food, gasoline, or health care.

This is important, because tweaking the value of the dollar might have radically different effects, depending upon which sectors get tweaked. Moreover, it's worth pointing out that inflation actually benefits debters (since the amount they have to pay back is worth less in the world than it was when they borrowed it) while hurting lenders (since the interest they recoup today buys less than it would have 10 years ago, when they made the initial loan or bought the bond or whatever).

When we talk about inflation, then, it makes sense to ask "whose inflation?" But even then, it's hard to see how we could ever produce a personally-applicable picture of the value of our currency.

posted by Skay @ 12:01 AM