Thursday, January 18, 2007

Fact checking

I have deadlines pressing and, as a result, I have been putting in long hours. So I decided to take a break and peruse the New Yorker that came the other day, the January 22 issue. The magazine has a well-deserved reputation for meticulous editing, even if they make odd style choices (there’s that “vender” again, I notice).

Numbers, though, are problematic. A great many people come to the writing and editing game having managed to indulge their aversion for all matters mathematical or scientific. They’re great with words—their passion—but having once decided they were math-impaired (whether that’s true or they just suffered abysmal elementary instruction in the subject), they continue through life not worrying too much if they get the fact right where numbers are concerned.

Don’t get me wrong. Fact checking is a big deal at the New Yorker. They’re quite careful with names and dates and places and quotations. But…well, let me quote the paragraph that prompted this post:
This year, Americans will consume close to four trillion kilowatt hours of electricity. In addition, we will burn through a hundred and forty-three billion gallons of gasoline, which at current retail prices will cost us some three hundred and sixty billion dollars, and twenty-six billion gallons of jet fuel, worth fifty billlion dollars. To heat our homes and businesses this winter, we will purchase sixty-two billion dollars’ worth of natural gas and heating oil, and just to grill our weenies we will buy some seven hundred and seventy-one million dollars’ worth of charcoal briquettes. In 2007, total energy expenditures in the U.S. will come to more than a quadrillion dollars, or roughly a tenth of the country’s gross domestic product.
Where to start?

Well, let me start with a copyediting nit. Where I come from, the word and does not belong in expressions like “a hundred and forty-three billion.” However, at least and was used consistently. So I suppose that’s a regional difference and I should shut up about that.

But let’s look at the numbers. That quadrillion got my attention. So I looked up the gross domestic product (one quick check on Google, no trip to the library required). It turns out that our GDP is on the order of twelve or thirteen trillion dollars. A quadrillion is a thousand trillions. A tenth of our GDP is somewhat above a trillion dollars, not a thousand times as much. But what’s a few zeros between friends?

Let’s see if the numbers add up correctly, even if we knock off the extra zeros.
“Four trillion kilowatt hours of electricity”
at roughly $0.10 each
Natural gas and heating oil$62,000,000,000
Charcoal briquettes$71,000,000
I’ll grant that the charcoal briquettes were thrown in as a joke, and somehow the diesel fuel for trucks, trains, and ships was left out. A quick Google check on diesel consumption, though, only adds another seventy or eighty billion dollars. So we’re really not very close to a trillion, let alone a quadrillion, and we’re well under a tenth of GDP.

Having gone through that exercise, I’m not sure I want to spend the time to read the rest of the article, nor am I likely to give the magazine as much credence in the future as I have in the past.

What’s the lesson? The lesson is that if you aren’t sure of your facts and you don’t know how to check your facts, you need to let someone check them who is competent to do so. Otherwise, your own credibility suffers.

Words and numbers matter.


ORION said...

I found your blog from Miss Snark! So interesting.
I am really glad I did!
Re: facts.
I am astounded at how detailed my editor got when she was going over my manuscript. Facts really are important and when you make a mistake it can come back and bite you in the you know what!

Dick Margulis said...

Thanks for dropping by!

Sounds like your editor was a pro. You lucked out—with your editor and your boat.

Dave Fragments said...

I don't know what point the article was trying to make. Yearly energy consumption figures are big.

All of those figures are contained in the Department of Energy's Annual Energy Outlook. These aren't secret numbers. the Energy Information Agency publishes the numbers every month, year, etc...

I used to work for DOE. I'm retired. If it weren't for "the Mouse that Roared being on TV, I wouldn't do this. ;)

DOE reports that Total retail sales of electricity in 2005 were 3,661 billion kilowatthours.

That could be close to 4 trillion if you wanted to really use rough estimates {?} NOT!!!

Also, DOE reports that
3,343,131,000 Barrels of U.S. Finished Motor Gasoline Product Supplied to USA in 2005. These are 42 gallon barrels not 55 gallon drums. So estimating - 3.3 times 42 is 138,000,000,000

That's not 43 billion.

What a mess. WHoever did that article needs a day or two to read the reports and put the facts together. I helped do that once and it took several of us a week complete with duplicate verification of the numbers and math.
Hard work...

Dick Margulis said...


The article was a profile of Amory Lovins by Elizabeth Kolbert. The paragraph in question was Kolbert's (I wouldn't want to blame Lovins for being that far off with his figures).

Thanks for correcting the figures I didn't take time to check. As you said, what a mess! However, the gasoline figure wasn't really off. She had 143 billion, not 43 billion. You have 138 billion. The two numbers are close enough for gummint work.

As for the electricity figure, what's a "kilo" between friends? Did you catch the YouTube recording of the guy arguing with Verizon over the difference between 0.02 cents per kilobyte and 0.02 dollars per kilobyte? It's a classic!

Dave Fragments said...

I'm on dialup so I don't do youtube. There is a big difference between 0.02 cents and 0.02 dollars (like 2 cents verus 2 mils, think property tax rates)

And you see what happened on 143 and 43 - I read it too fast and missed the "one hundred and" ... THat's why numbers are so hard and why you must have help and fact checking.

As for kilo - A kilo is 1000 of anything. A kilometer is one thousand meters or 1000 meters.

An average lightbulb, let's use 60 watts consumes 60 watt-hours of electricity in one hour. A kilowatt-hour (they always leave the hour off, everyone knows it's there, oh wait, the science types know its there).

A kilowatt is 1000 watt-hours of energy. It's the equivalent of burning the light bulb for 1000/60 or 16.6 hours (give or take a little bit).

My background is chemical engineering - BTW

I should add one thing, the EIA and the rest of the world separate energy consumption into four categories: Transportation, Residential heating, Electricity consumption and Industrial use.

Can you see why?

It's simple - Transportation (motor fuels, aviation fuels) are made from petroleum.
Electricity is mostly made from coal.
Residential heating is normally natural gas (methane).
And Industrial uses are predictable loads regardless of the fuel source. They are also non-essential in winter or summer peak loads crises.

Dick Margulis said...


Yes, we're on the same wavelength. My point about a "kilo" was exactly that it got lost between watt hours and kilowatt hours in the New Yorker article.

Sorry you're on dial-up. You'd have enjoyed the recording. Basically, the guy was quoted a price of 0.02 cents and got a bill at the 0.02 dollars rate, and nobody at Verizon could comprehend the difference. All they could do was punch ".02" into their calculators and multiply, with no regard for units or where the decimal actually belonged.

Zen of Writing said...

I'm with you on the words and numbers issue. I did corporate editing for a bit, and had to pay attention to numbers and mathematical formulas. I often had the feeling that the people reading the reports wouldn't have noticed, however. It was my own personal purgatory.

Now I have a different job, and I feel much better.

Nice blog, also found you from Miss Snark.

Anonymous said...

checking facts -
Your blog is posted 01/18/07 but you reference the 01/22/07 New Yorker.

Dick Margulis said...


You are correct. Magazines are identified by the cover date, which is generally some days after subscribers receive the magazine in the mail.

Anonymous said...

Quadrillion?! That number doth indeed reeketh of snake oil.

As an erstwhile student of resource management, I may have an insight into where such apparent inflation creeps in.

For some economic purposes, the value of a resource is counted not as the present value of the resource in the marketplace today, but as something far more esoteric. It could be the summed value all of the money that changes hands along the way from in-ground to final consumer. Or worse, it could be based on a scenario, like "What if OPEC throws us under the bus twenty years frm now?" Or even more provocative, "What would the value of the last barrel of oil be if the mid-east oil fields suddenly are revealed to have gone dry fifty years ahead of schedule?"

When it comes to statistics, especially gov-a-munt statistics, there is always the chance that someone, who will not be accountable, has skewed them for some special purpose. I think we saw a similar force at work in the report on global warming a few years ago.

Thanks for blogging their fact unchecked mis-step, Dick. Whatever accounts for their error (maybe Murdoch bought them in a secret deal?) I'll be watching more closely what I read there.