Numbers can look so real...

Few things are more comforting to our rational minds than neat rows of numbers and logical calculations. Because the math works, it must mean the human activity it describes works as well.

But what if the numbers are...well, semi-goofy? As we rely more and more on computer analysis to plan our actions and guide our choices, we might want to pause and examine the original data stream.

For example, if inflation is reportedly pretty tame why does it feel so much worse in my wallet? Bill Gross, the manager of Pimco Total Return, the nation's largest bond fund, refers to the CPI as a "con job" that deliberately understates the price pressures faced by Americans in order to keep Social Security payments and other government costs pegged to the index unduly low.

In a report about the CPI, he noted that some of the adjustments don't accurately reflect how much consumers pay for goods. Pimco estimates that the changes have shaved more than a percentage point off the CPI.

"Did your new model computer come with a 25% discount from last year's price?" Gross wrote. "Probably not. What is likely is that you paid about the same price for memory improvements you'll never use."

Another flaw with the CPI numbers is that the government now assumes that higher prices for one item will lead consumers to buy more of a substitution item. That may be true. But if people buy fewer steaks and more hamburgers, for example, it's unrealistic to say that inflation isn't a problem, skeptics maintain.

"The government can claim there's no inflation but all they're measuring is a reduced standard of living," argues Peter Schiff, president of Euro Pacific Capital, an investment firm specializing in overseas investments. [More]

The disparity between policy numbers and personal impact helps breed, I think, a contempt for rational economic policy decisions by government. It also produces rude surprises when consumers react differently than [inaccurate] numbers predict, making policy decisions more tentative in the aftermath.

Another area is the current intense scrutiny laboring in a blizzard of figures is energy, and more specifically, oil. Again, while the math is elegant and illuminating, the input is suspect.

China, which has grown into the world's second-biggest oil consumer after the U.S., stands out as a particular problem. Just ask Eduardo Lopez, who tries to dope out the China market as the senior demand analyst for the Paris-based International Energy Agency, an affiliate of the OECD. He says China does not report demand, leaving him and others to figure it out from data on production, trade, and inventories. What's more, he says, "there are thousands of so-called teapot refineries all over China" that are technically illegal and therefore left out of China's official statistics.

Making his job even more trying, China appears to be creating a strategic stockpile of oil, but has never acknowledged it, Lopez says. If Lopez and others are underestimating how much oil China is squirreling away, then they're inadvertently overestimating true global consumption, and vice versa if they've overestimated China's stockpiling. [More of a must-read]

Without clearer information, I think we can brace ourselves for strange market responses. Also because official information is so suspect in many countries, private data commands a huge premium and enables those who know to profit mightily. This is unfortunate as it undermines the working of an efficient market and produce its own side market such as exists around Warren Buffet. The actions of those with purportedly inside info will be mimicked as quickly as possible.

We are not exempt here in agriculture. Although I lean more toward incompetence than intrigue, USDA figures strike some (like our own Chip Flory) as occasionally fishy.

There is a modern weakness for linking accuracy with precise numbers, even though the number is a figment. Call it a numerical prejudice, but our efforts to know things "exactly" probably would be better spent understanding the underlying causes and effects.