Data Is Data. Except When It Isn’t.
One of the most frustrating aspects of Science-with-a-capital-S is the certainty that data always comes with noise. That is, for one reason or another, to one degree or another, the data you deal with is bad. Sooner or later a judgement must me made, and some of the information is removed because of the judgement.
For that last sentence, read: A scientist always has the opportunity to impose his beliefs on the data, and will do so.
Surprisingly, the situation is even worse than that. If you believe that an instrument that was created to make a measurement can by definition make no judgement about the data, you are quite mistaken. The instrument can only detect the events that it was designed to measure, and can only do that within the limitations of its design. The very act of choosing that instrument, then, determines what the scientist will “see”. What the data tell him has been pre-determined by the selection of the instrument.
If you read my scribbles long enough, you’ll know that I’m not particularly impressed with scientists who insist that science is “objective” (or that they are, for that matter). They understand not at all the nature of understanding, much less the nature of their own prejudices.
So let’s talk about oil. How much is there, anyway? Are we about to run out? Seems a simple thing to measure. How much do we get when we add it up, and how much do we consume?
From Yves Smith at Slate:
Think of the host of financial and economic statistics you see every day: stock prices, unemployment data, company results, foreign exchange prices, GDP growth rates. Market information is precise. Government statistics, while a little fuzzier, are complied and computed in a generally consistent fashion.
But oil is in a completely different category: It’s a strategic resource bought and sold internationally. Many countries, either by indifference or design, simply don’t provide reliable information.
By indifference or design. Say you were assigned to measure the output of an oil refinery. Say you were dependent on the paycheck that came every week, signed by the governor, whose party insisted that more drilling was unnecessary. Or, whose party insisted that there was plenty of oil just off shore. Which way does your boss want you to “round” the numbers? You know you can’t trust the numbers that the refinery has provided, but you can (and probably must) use them. But to what end? What set of numbers, the weekly (which showed a decline last week) or the monthly (which didn’t) are you going to list first in the report? – the report that will either be released to the AP news wire, or to the drillers union trade journal…
Now say it’s 1994, and your “governor” is Saddam Hussein. Is there any objectivity to be found here at all?
The International Energy Agency, a premiere source of oil-related information, was caught off guard by the surge in oil prices, so it decided it needed to get a better grip on capacity. The IEA is only partway through a survey of the world’s biggest oil fields, yet says it expects to show a significant reduction in estimated reserves.
But even when concluded, this project will be far from reliable. For starters, take OPEC, which is estimated to control two-thirds of the oil reserves and to provide 36 percent of oil production and is largely unresponsive with IEA inquiries. And cooperation doesn’t always translate into insight. For instance, Iraq recently claimed it has as much as 350 billion barrels of oil, triple its proven reserves and more than even oil kingpin Saudi Arabia has. Is this claim completely crazy? No one knows for sure; Iraq is underexplored, with only 2,000 oil wells versus more than 1 million in Texas.
We can play this game with global warming and temperature data too.