We’re Flying Blind Time for New Economic Metrics

Hal Plotkin, Special to SF Gate
Tuesday, June 22, 1999

URL: http://www.sfgate.com/cgi-bin/article.cgi?file=/technology/archive/1999/06/22/metrics.dtl

Garbage in, garbage out.

The familiar dictum from the computer industry, known as GIGO, says that if your initial data is faulty, your final conclusions will also come up short, no matter how mighty your computer. Thanks to the technology revolution, GIGO is spreading like a computer virus. Government policy makers have caught the bug big time.

The latest example was the conflicting testimony offered two weeks ago by Federal Reserve Chairman Alan Greenspan and Microsoft founder Bill Gates to a congressional panel looking into the impact of technology on our economy. Greenspan warned Congress we might be nearing the end of the period when technology fosters the kind of increases in productivity that has spurred growth in recent years.

Greenspan was one of the first to talk about technology’s positive impact on productivity, so his recent backtracking is worrisome. Gates, meanwhile, squeaked out the opposing view, maintaining that we are just beginning to see the impact of technology on productivity.

On this topic, I side with Gates. But what is far more interesting is that right now, due to antiquated Federal economic statistics, not even our best minds can point to much in the way of solid facts or data to back up their opinions. Instead, like the old story about blind people describing different parts of an elephant, even experts must rely on their own limited senses and differing perspectives to formulate guesses about what is happening.

Despite costly Federal programs that collect and analyze economic data, economists are essentially flying blind. As a result, many of our key policy makers are clueless about current real-world economic trends.

Take, for example, three critical economic measurements, the aforementioned productivity numbers, the consumer price index, and most particularly, the Federal poverty index.

How do you measure technology-driven increases in productivity? If you are the U.S. government, not terribly well. According to Federal figures, U.S. productivity rose at an annual rate of 2.7 percent per year between 1947 and 1973. Since then, the Feds say, productivity has been increasing at less than half that rate, just 1.1 percent per year, on average. These more recent productivity figures, many now agree, are completely ludicrous. Fortunately, Greenspan ignored them when formulating the Fed’s policies on interest rates.

As Greenspan understood, productivity is dramatically, perhaps immeasurably, impacted by technology. Spell-checker software programs alone, for example, have probably contributed at least a 1.1 percent gross productivity increase to our economic output. Unfortunately, government statisticians have no way of making meaningful comparisons between, say, a 1970’s-style typing pool and a single word processor.

In fact, the Feds are just now catching up to the fact that the burgeoning information economy even exists. It was just two years ago, for example, that the North American Industry Classification System (NAICS), a key tool used to measure economic output, was amended to include a new category called the Information Sector. Welcome to cyberspace, Uncle Sam. We’ve been expecting you.

The same obliviousness infects Federal measures of inflation, currently said to be running at just 2.6 percent, up one percentage point from last year. But if you look at inflation figures a little more closely, you discover that economists from both political parties have been cooking the numbers.

Back in the early days of the Reagan administration, for example, when inflation figures were running unacceptably high, the Feds removed the cost of purchasing housing from official inflation calculations. Instead, they devised a complicated formula that theoretically describes how much people might need to pay for housing they don’t own.

The Reagan administration decided that if the real price of something, in this case, a house, is going up too fast to keep inflation figures low, you just remove it from the index. Call it “creative accounting.”

Similar sleight of hand has taken place during the Clinton administration, also under the guise of improving the utility of economic statistics. A series of “methodological” changes in the way the consumer price index is calculated were implemented during Clinton’s watch.

The advocates of this approach say it better reflects how people really live; if the price of steak goes up to $10 a pound, more people buy ground beef. So you stop measuring the price of steak and instead measure the price of the ground beef most people actually buy. Never mind that they might prefer steak. The quality of the stuff on their dinner plate may have gone down but, voila, inflation has remained low. Re-elect us! Stay the course! Times have never been better!

And, to be fair, these are better times. At least for some of us. But not for people living at or anywhere near the Federal poverty level, another example of an antiquated and even dangerous economic statistic. Did you ever wonder why people living above the Federal poverty level still live in poverty? According to the Feds, it takes just $8,240 per year for one person, or $11,060 for two people, to escape poverty.

The problem is the Federal poverty index was established back in 1964 with the assumption that food costs would average one-third of all expenses. But in the intervening years, food budgets have slipped relative to other needs, most notably housing, which is now significantly underweighted in poverty calculations. According to the Feds, a single person making just $8,240 per year is able to escape poverty because he or she only needs to pay somewhere around $250 per month in rent. Yeah, right.

There is no way we can fix these faulty economic statistics with further politically motivated tinkering. Leaders of the National Association for Business Economics have recommended scrapping the entire cumbersome, discombobulated Federal economic data gathering apparatus

and bringing it all under a single agency, led by a Federal economic statistics czar.

I think we should go even further. We need to call a sit-down meeting with some creative thinkers to figure out an entirely new set of metrics that can better measure the economy that now exists. It’s OK with me if we invite some economists to the meeting, but we also need sociologists, anthropologists, psychologists, historians, and others. Maybe even a few lowly business writers. We can ask ourselves, what should we really be measuring?

If we can’t accurately measure the impact of technology on productivity right now, for example, perhaps we should put that on a backburner and look for a different set of leading economic indicators. At least that way we would not be burdened with potentially misleading economic statistics

What might these new economic indicators look like? We could start by counting the number of people who want, but don’t have, a place to live. We could count the number of people who want, but can’t obtain, better educational opportunities, health care insurance, or reliable transportation. We could count the number of 50-year olds who are forced to accept lower pay in new jobs. We could track suicide rates in different age groups. We could measure the amount of parenting time available in both one-and two-parent families.

I can’t say, all by my lonesome, exactly which items we should be measuring to obtain a more accurate sense of our economic health. But it is painfully clear that our current economic measurements are unsuited to the technology-driven times in which we live. It’s unlikely we’ll ever come up with the right answers until we start asking more of the right questions.

In the meantime, though, as a reporter covering technology, I’d like to let the Fed chairman know that, when it comes to the impact of technology on productivity, we ain’t seen nothin’ yet.

CC BY 4.0
This work is licensed under a Creative Commons Attribution 4.0 International License.