Richard Tol’s Many Mistakes Prove Him Right

This blog was created because I believe the world is insane. I can think of no better example than people saying up is down. Or rather, Richard Tol saying work finding global warming will be dangerous indicates global warming is not as bad as we thought.

I’ve written a number of posts mentioning Richard Tol lately, but I’ve intentionally avoided discussing the merits of his work. I can’t do that any longer. Tol has just published a correction to his work in response to his critics, and it’s too funny to ignore.

For some background, when Tol attempted to model economic impacts of global warming, he flipped the sign of a couple data points. He showed estimates which showed global warming will be harmful as showing it will be beneficial. He claims this happened, I kid you not, because of gremlins. I guess that’s one way to avoid taking the blame.

More amusing is the fact when Tol corrected the mistakes, he promptly said global warming isn’t as bad as we used to think:

A careless reader could easily take this to mean fixing his mistakes causes the results to be less scary, meaning changing data to be negative instead of positive makes our results go up. Up is down.

Of course, that’s not really what Tol says. Tol shows the effect of the corrections is to shift all of his results downward. We can see that in this figure:

5-5-Tol_Correction_Fig_1

We can see the actual reason things are less bad (according to Tol) is new data was added:

5-5-Tol_Correction_Fig_2

So up isn’t actually down. You might be wondering why I’m writing this then. New data, new results. Who cares, right? Wrong. Look at the two images above. Notice how the x-axis is different in them. The first image only goes to three degrees. The second image goes to 5.5. What do you think would happen if we showed Tol’s original model on the same scale as the second image?

Let’s check. Here is the original model (black) and the new model (red) on the same scale:

5-5-Tol_Correction_Display

Now it’s easy to see why Tol’s new results say things aren’t as bad. Practically nothing could look as bad as the old results. It doesn’t matter that a positive value was changed to negative. It doesn’t matter that the new results are all negative:

3.0	-1.1
3.0	-2.5
3.2	-11.5
1.9	-0.5
2.9	-1.8
5.4	-4.6
2.9	-2.0

It doesn’t matter all the new details indicate things are bad. It doesn’t matter none of the new details indicate global warming will be beneficial. Nothing could possibly be as scary as Tol’s original model made things out to be.


None of the new details indicate global warming will be beneficial. Tol admits the new results show “impacts are always negative.” How then can things be less bad than before? Tol wrote in the recent IPCC Report:

Estimates agree on the size of the impact (small relative to economic growth) but disagree on the sign (Figure 10-1). Climate change may be beneficial for moderate climate change but turn negative for greater warming.

But now we see the estimates in general don’t disagree on the sign, and climate change won’t ever be beneficial. It seems every positive thing Richard Tol said about global warming was wrong, yet somehow, he claims that proves he was even more right than he had thought.

And while those positive things he said were important enough he modified the IPCC report to add them in, he now tells us:

The full text states that the initial impacts of climate change are irrelevant for policy making, as they cannot be avoided. It therefore does not matter whether they are negative or positive, big or small — the case for greenhouse gas emission reduction is the same.

In other words, a topic Tol has repeatedly discussed, in the media and in his published work, doesn’t matter. A topic Tol went out of his way to make sure got included in the latest IPCC report doesn’t matter. It doesn’t matter no data shows global warming will be beneficial. It doesn’t matter all the new data shows global warming is dangerous. Things are still better than we thought.

You have to admire the audacity of Tol’s argument. He never showed his original model’s catastrophic predictions. He only showed its happy predictions. Now, when those happy predictions turn out to be wrong, he says it doesn’t matter because the catastrophic predictions he never talked about are wrong.

I’m not sure how responding to criticisms of his mistakes by pointing out he made even more mistakes is supposed to make things better. Maybe the gremlins can explain.

I should point out I noticed Tol’s published correction lists two estimates from Nordhaus 2006, one for a temperature change of 2.5 degrees, another for a change of 3.0 degrees. This is odd because the data file he published lists them as both being for 3.0 degrees. The file even explicitly highlights the fact he changed the one value from from 2.5 to 3.0. It appears Tol made a mistake he forgot to acknowledge and correct.

Advertisements

One comment

  1. I just came across an incredible point regarding this issue. Back in 2009, Julie Nelson commented on a paper by Richard Tol, saying:

    (One of these estimates, from Hope 2006, is described in Tol’s footnotes as the estimate used in the Stern Review; however, there must be a mistake here: the projection of significant economic gains at 2.5 degrees of warming, shown for that study, is clearly at odds with Stern’s message and conclusions.)

    Tol responded to her, yet he somehow failed to address this point. Years later, he published this correction, proving she had been right all along.

    For something like half a decade, Tol knew (or should have known) he inverted the sign of at least one of his data points, and he did nothing.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s