The future might not be like the past

by Russ Roberts on February 14, 2007

in Data

Public discourse on global warming reached a new low recently with this assessment by Ellen Goodman:

By every measure, the U N ‘s Intergovernmental Panel on Climate
Change raises the level of alarm. The fact of global warming is
"unequivocal." The certainty of the human role is now somewhere over 90
percent. Which is about as certain as scientists ever get.

would like to say we’re at a point where global warming is impossible
to deny. Let’s just say that global warming deniers are now on a par
with Holocaust deniers, though one denies the past and the other denies
the present and future.

Where to begin in analyzing that last sentence? The moral vulgarity of it? Or just the scientific inaccuracy? True, Holocaust deniers deny the past. Global warming deniers are denying the reliability of statistical models of the past being used to forecast the future. My understanding of the science is that the models of the past aren’t quite theory-free but they’re not exactly theory-robust, either. That is, we don’t fully understand the causal relationship between human activity and global temperature, so much of what is being predicted about the past is an extrapolation.

(Mark Steyn, on the other hand, knows where to begin. He is brilliant. HT: CHCH, Don Boudreaux)

Gregg Easterbrook at ESPN (and the author of The Paradox of Progress) has a very nice summary of bad predictions made in the past year, reminding us that in many areas of life, the future is not like the past. Most of them are sports predictions, but he throws in some nice gems from outside the sports world:

Bad Hurricane Predictions: The year of Katrina and Rita,
2005, obviously was awful for Atlantic cyclones, with a record 15
hurricanes. Both the National Oceanographic and Atmospheric
Administration, and media-favorite hurricane forecaster William Gray of
Colorado State University, predicted 2006 would be bad, too. In
December 2005, Gray predicted for 2006 nine hurricanes, five of them
intense; there was an 81 percent chance a major hurricane would strike
land in the United States in 2006, Gray and CSU said with ridiculous
pseudo-precision. In May 2006, NOAA forecast eight to 10 hurricanes,
six of them intense. The 2006 hurricane season would be "hyperactive,"
NOAA declared: "The main uncertainty is not whether the season will be
above normal but how much above normal it will be." Actual: In 2006
there were five Atlantic hurricanes, two of them intense — pretty much
smack on the 20th century average for the Atlantic basin. None made
landfall in the United States.

Easterbrook also gives this beautiful example of a really bad prediction from 1986:

In June 1986, Newsweek ran its infamous "Marriage Crunch" cover story,
which pronounced that a college-educated career woman of age 30 had
only a one-in-five chance of winning a husband, while an educated
professional woman of age 40 had essentially no chance. At a time when
this must have sounded funny to someone at Newsweek, the magazine
declared that a single 40-year-old career woman was "more likely to be
killed by a terrorist" than to find a man who would say "I do." Twenty
years later in June 2006, Jeffrey Zaslow of the Wall Street Journal
checked to see how these predictions stood the test of time. Women aged
30 to 40 in 1986 when Newsweek declared them unmarriageable are aged 50
to 60 now, Zaslow reasoned. Crunching Census Bureau stats, he found
that 90 percent of college-educated American women between the ages of
50 to 60 have married at least once. Zaslow tracked down the 10
career-shark single women who were named in the 1986 Newsweek cover
story as certain spinsters: eight later married.

Be Sociable, Share!



28 comments    Share Share    Print    Email


Dale gribble February 14, 2007 at 5:50 pm

the global warming chicken littles are themselves history deniers. anyone remember the coming ice age after we had a few cold winters in the 70's? Or better yet whatever happened to the "nuclear winter" Prez Reagan would bring on by standing up to the Commies? Or will a little nuke exchange with North Korea cure our global warming problem?

Tim V February 14, 2007 at 6:13 pm

If 90% certainty makes something "impossible to deny", why does my econometrics teacher insist that 95% confidence is the going norm?

rjh February 14, 2007 at 6:35 pm

If you read Gray's actual forecasts at you will find that they are very precise descriptions of technique, accuracy estimates, etc. It's the PR people and news reporters that transform excellent scientific results into the news nonsense.

Mesa EconoGuy February 14, 2007 at 7:14 pm

While we’re on bad sports predictions, nobody has brought this one up in a while: Craig James, then CBS college football analyst said in 1997, “There is no controversy – Peyton Manning will win the Heisman Trophy.” Oops.

And here’s one from the financial world. Bill Gross, a.k.a. Sonny Bono, predicted Dow 5000 in 2002:

He only missed by about 2100 points.

The difference between the global warming nuts and these bad guesses is that they’re basing a course of forced action on their bad guesses, while these stinkers were absorbed in stride by free participation.

Wes Fontana February 14, 2007 at 7:33 pm

LOL at the ESPN article. Check out the second catagory about Miss American- proof that the world is getting better!

Person February 14, 2007 at 11:49 pm

Russell_Roberts: I have to ask: why did you regard an Ellen Goodman column as worthy of a response? I'd like to know.

kurt February 15, 2007 at 10:22 am

I thought Bayesian inference is still controversial? How can you speak of a consensus on climate change then?

Matt C. February 15, 2007 at 10:47 am

James Taranto First pointed this out last friday. He makes a very convincing argument against Goodman's statement.

Matt C. February 15, 2007 at 10:47 am

James Taranto First pointed this out last friday. He makes a very convincing argument against Goodman's statement.

Bruce Hall February 15, 2007 at 1:04 pm

One of the interesting aspects of global warming debate is the significance of climate change. No one really questions the data as much as they use charged adjectives to buttress their opinions.

I have begun to look at the issues in a slightly different way: it is not the data, nor the words that create a gulf of understanding… it is the way data are presented.

The first of three (possibly more) installments was posted today:

For additional information about this issue, you might find this from Climate Science's Dr. Roger Pielke, Sr., at Colorado University, interesting:

Ray February 15, 2007 at 2:04 pm

The solution is to not read economists' comments about science and not read scientists' comments about economics. Both use models and sources to demonstrate their biases.

Randy February 15, 2007 at 2:12 pm


Given my own concept of weather, I'd have to go with a graph using about a 20 degree range. Because unless the temperature changes by 20 degrees or so, I'm wearing the same jacket I wore yesterday. Indoors I'm a bit more choosy. That last scale would work for indoors.

Bruce Hall February 15, 2007 at 2:20 pm


Check the site tomorrow and you'll see you read my mind.

M. Hodak February 16, 2007 at 10:07 am

As far as I know, even discounting my own family's experience, the odds that the Holocaust actually happened approach 100%.

Previous post:

Next post: