How Well Can Economists Measure Variety?

Variety is a good gauge of progress. But how well can economists measure it?

The New York Times, "Economic Scene" , April 19, 2001

Are you better off than you were four years ago? It's not just a rhetorical jab for presidential candidates; it's a vexing puzzle for economists. From a purely material point of view, what does it mean to be better off? In a dynamic economy where goods and services are constantly changing, how do you gauge progress?

The Federal Reserve chairman, Alan Greenspan, raised the issue in a speech late last month, arguing that it's more important to improve economic measurement than to tweak forecasting models. Before economists can say that the price of a good has gone up or down, which is essential to measuring productivity or inflation, they must define what the good is. For a 12-ounce can of Diet Coke, the problem is trivial. For health care, it is not. Is the good a particular procedure or drug, the treatment of a given condition, or a particular outcome? Ubiquitous "new and improved" products further complicate the picture.

While never easy, the measurement problem is becoming increasingly difficult. A few decades ago, Mr. Greenspan noted, "an average price of hot-rolled steel sheet and a corresponding total tonnage was precise enough for most analytic needs."

"By the same token, tons of steel per work hour in a rolling mill yielded rough approximations of underlying productivity for most purposes," he added. But economic value increasingly comes from less generic, harder-to-measure and often intangible goods. To understand the economy and avoid misstating the rates of inflation or productivity growth, economists need to find better ways to measure what's actually going on.

That's a hard task, especially given the data available. Consider what most businesspeople believe to be one of the most important developments in recent years: an increase in product variety. "We boring economists are left to say, 'What hard evidence do we have of that?' " says Peter J. Klenow of the Federal Reserve Bank of Minneapolis. "You may be right, but how do we quantify that? What data can we use to systematically measure that?"

Even if variety is increasing, that doesn't prove that consumers are better off. Perhaps they're bypassing the proliferating new brands in favor of old standbys. In that case, variety would just be noise. And there's also the issue of exactly how much variety has grown. Is it really such a big deal? "The question is whether something grew 3 percent in a decade or 30 percent in a decade," said Mark Bils, an economist at the University of Rochester.

He and Dr. Klenow tackled the variety question in a paper presented at this year's American Economic Association meetings and scheduled for publication in The American Economic Review. (The paper, "The Acceleration in Variety Growth," is available here.) To do so, however, they had to be ingenious. That's because the government doesn't collect statistics on product introductions. To the contrary, the Bureau of Labor Statistics emphasizes old products. The bureau's agents go out each month and look for a list of the same products -- a certain weight box of Tide, for example -- and record their prices.

That's how we get the Consumer Price Index. Most of the time, the bureau adds new products only as substitutes for old ones that have disappeared.

This quirk of the data gave the two economists an idea. Perhaps they could measure the arrival of new products, which the government doesn't measure, by looking at the disappearance of old ones. If variety is increasing in a certain category, like breakfast cereals, the new products may be displacing old ones; there would, in short, be more product turnover than in other categories.

Alternatively, it's possible that products are disappearing because no one wants them. In that case, consumers would be spending less on those categories. Dr. Klenow and Dr. Bils divided the government's product categories into two groups, depending on turnover. "Dynamic" goods and services showed frequent changes, while "static" products almost never changed. Not surprisingly, various sorts of clothing were dynamic, while staples like milk and gasoline were static.

The economists then looked at consumer spending from 1959 to 1999 and found a striking result. Even after correcting for income growth, relative spending on dynamic goods increased 1.3 percent a year, "a very sizable shift of 68 percent over the past 40 years," they write. When they divided the time period in half, they found an even stronger shift. Seventy percent of the movement toward dynamic goods took place from 1979 to 1999. That movement appears to have accelerated in the mid-1980's, Dr. Klenow says.

He argues that the shift from static to dynamic goods indicates that consumers really like the new goods. To get them, people are willing to spend less on other things. Television used to be a static category until cable increased variety, and telephone service was stable before cell phones. Now both are dynamic categories, and they're attracting a bigger portion of consumers' budgets.

The same is true for more controversial shifts, like increased spending on prescription and nonprescription drugs -- both dynamic categories. Consumers are voting with their dollars in favor of more variety. "It's not like we'd be happier or better off if we just said, 'No, no new drugs. Only old drugs,' " Dr. Klenow says.

But these highly visible sectors have no monopoly on variety. Along with all sorts of consumer electronics, prescription and nonprescription drugs, and physician services, dynamic goods include such mundane categories as breakfast cereals, candy, sports equipment and toiletries and cosmetics.

There are not only more brands on the shelves (and more different services available), but people are spending more on the categories that offer them more choices. Increasing variety raises Americans' standard of living, by giving us more of what we actually want rather than forcing us to make one-size-fits-all choices.