Monday, August 1, 2011

We Should Listen to Krugman

We don't have time to do research projects on every topic, so probably most people live with imperfect short cuts when we want to figure out the answers to complicated or controversial questions.

One option is to try and figure out the consensus opinion and go with that. That's a good first order approximation. Another is to try and look at track records. That plays a big role for me.

The right really turned me off for the first time with the Iraq WMD failures. Not that the left didn't think Saddam had WMD. But the right was loudly touting conclusions based on bad evidence. The experts on the right in my view had a preferred conclusion and it overrode their ability to evaluate the evidence correctly.

The right did the same when it came to the motives of Islamic terrorists. It's obvious to me that Islamic terrorists are motivated by US violence. But on the right the view that they hate us because we're so great, which is obviously a preferred view since it strokes the ego, has a major foothold. The right has a powerful ability to allow their preferred conclusions to blind them to what in my view are obvious errors.

When Reagan raised taxes the right predicted dire economic consequences. The same happened when Clinton raised taxes. Those predictions were quite wrong. The right predicted that the Bush tax cuts would lead to prosperity and growth. What we actually have is just about the worst economic performance since the Great Depression in the wake of Bush's economic policies.

These things matter a great deal to me. It impresses me when a study shows that Paul Krugman is the top prognosticator, and Cal Thomas the worst. These things are not easy and you wouldn't expect even the smartest person to be right all the time. You can't anticipate some things that can effect predictions.

But probably his evaluation of this debt ceiling limit deal is worth considering. The Republicans pretty much got everything they wanted. No tax increases, massive cuts in entitlement spending. This in the middle of a terrible economic down turn with 9% unemployment. Here's Krugman on This Week(watch Part 2 here). He's expecting continued extremely high unemployment through the end of next year. He doesn't add this, but I suppose that's not going to bode well for Obama's re-election.

And I tend to think that's a good thing. The guy right flanks even Republican voters. That's been true for war, health care, and now taxes/entitlement spending. And lest you think this is about Obama being just unable to overcome Republican pressure, read Greenwald. Same old thing. If there's no price to be paid for a Democratic President when he acts like an extremist Republican, then why shouldn't everybody act like an extremist Republican.


HispanicPundit said...

You write, The experts on the right in my view had a preferred conclusion and it overrode their ability to evaluate the evidence correctly.

The same could be said of Krugman. Megan McArdle has blogged on this here and here.

For more on Krugman's predictions in general, see here, here, and here.

Let me just make clear though, this is not to say that I disagree with your conclusion. Of course I think Paul Krugman's predictions will be better than his peers...but why? That brings me to my next point.

You write, It impresses me when a study shows that Paul Krugman is the top prognosticator, and Cal Thomas the worst.

This is one of the things that irritates me about you Jon. You don't compare apples to apples. You seem to like to stack the cards against conservatives - it's as if you wanted to be a liberal, no matter what the data shows.

Do you see anything odd with your comparisons above? Click on the link. Notice that Krugman is the ONLY economist out of all of them. A nobel prize winner at that. You don't think that a trained economist, regardless of political party, will have a higher prediction rate on issues related to the economy than non-trained pundits??? Would it be fair for me to say, "Bernard Lewis knows more about the middle east than Jon" therefore Bernard Lewis is more likely to be right? Cuz that is what you are doing here.

I mean, Krugman has training in the very thing he is predicting about. Of course he is going to be more right. This speaks more about the science of economics, than anything else. This is why I read ALL economists, left and right. Because even the leftist economists, as long as they have economic training, will be miles (and I do mean MILES) ahead of your typical leftist pundit. Trained means something.

A real comparison would be Milton Friedman to Paul Krugman. Or atleast an Edward Glaeser or a Martin Feldstein. But you probably can't do that, cuz we both know you don't read them. You prefer to compare the liberal giants, to the conservative pundits - and you refer to that as 'giving both sides equal say'. Sigh.

Jon said...

A couple of things here. First of all, be careful to not misread me. The title is "We Should Listen to Krugman" not "Krugman is Always Right." You and I agree on this, and I know you read Krugman and regard his thoughts as worth considering. That's the main point.

What's floating in my mind here really is how so many of the people I know don't seem to be affected by colossal and destructive predictive failures, like the WMD failures. Did you know that VD Hansen was like Cheney's hero. If I recall I think Cheney used to get together with people and it was like VD Hansen book club. Cheney's subsequent errors, and the resultant destructive policy, was a direct result of his reliance on Hansen. That matters to me.

I'm not comparing Cal Thomas and Paul Krugman. I just copied the title of the article. The study is not about economic predictions. It's about predictions of the punditry. The study quotes Glenn Greenwald and others that talk about how our opinions are shaped by pundits and there's no accountability for errors. That's kind of what I'm talking about here. How are opinions about complicated and controversial subjects shaped? Pundits are regarded as experts, but we need to look back on their track record. That's my belief. So here's a study. You can criticize the methodology. You can say that yeah, over the period of a few decades you can find errors from the top performer here. But the performance matters to me and I agree with Greenwald. We should be looking at their records.

I didn't run the study so I don't get to choose who is involved. Yeah, I agree with you that it would be great to compare economists. But that's just not what this study was trying to do. I think it's still worth looking at.

Jon said...

But let's take a look at these links of yours that show Krugman's predictions are not accurate. Looking to this history is something we should be doing more in my opinion.

The first is Megan backpeddling because she was wrong about the 2008 recession and Krugman was right. Her response is that Krugman has been doom and gloom forever, so if you constantly predict doom and gloom of course it will happen eventually. But she doesn't actually show that he did this. The only hard evidence at the first link is the evidence that shows Krugman is right and Megan is wrong.

Megan's second link takes us to Mankiw who shows that Krugman was incorrectly rosy prior to the 2001 down turn. So there's data showing Megan is wrong. He's not doom and gloom all the time. Krugman isn't really saying "We're not going into a recession." What he says is the data don't support gloomy economic views, so Bush is just trying to scare people about the economy in order to ram through a tax cut. Is he technically sticking his neck out and making a prediction? I read it as more tentative then that, but he's certainly not predicting a down turn, so if you want to call that a strike against him that's fine.

Links 3 and 4 are very interesting. Read them in tandem. In the first Megan says Krugman ought to be embarassed because he called for a housing bubble in 2002 to stimulate demand and pull us out of the recession. So then read the next one from Megan. She links to Krugman in 2005. He says yeah, the Fed did exactly what he suggested. He created a housing bubble and it worked. But it's gone on too long. There was room to inflate housing prices in 2002, but here we are in 2005 and it's out of control. We've used the room and we're creating bad conditions. This could be really bad. My goodness, if that's poor accuracy what would good accuracy look like?

True enough the remainder of the fourth link from Megan goes way back to ping him on errors regarding the internet, predictions of a raw material crunch, productivity growth. Fair enough. Those are errors.

In your fifth link Megan comments on the criticism she offered in the third link, so there's not much to that.

HispanicPundit said...

Regarding Krugman being wrong: Megan McArdle's primary audience are long time bloggers. She hardly posts links. But long time readers of her blog know its true - everybody reading Krugman has seen him as doom and gloom for a long time. I have known this. Nobody really questions that premise of hers.

With that said - you do agree that ALL predictions must be taken into account - both, the positive AND the negative ones, right?

Regarding WMD: why blame Bush and Cheney for that? The blame for that should be the CIA. Remember, Richard Clark, someone who you have quoted in the past in support of your views, said "we all believed Saddam had WMD". Those are strong words. He didn't suffer from the "preferred conclusion" psychology you refer to and yet was equally fooled.

If there is one lesson we learned from the Iraq war, it's that we should take the CIA's assurances with a lot more caution. Anything more than that, is a stretch, IMO.

Jon said...

On Krugman, you could be right. I haven't been reading him for a long time.

I definitely agree that we should look at all predictions, good and bad. That's key for me.

On the CIA, take a closer look at your Clarke interview. He says "We all believed." Meaning who? In context he means the administration, not necessarily the CIA. The very next sentence he talks about the cherry picking of data done by Rumsfeld and Wolfowitz. They didn't handle the data the way trained analysts do. This is another Team B situation. They didn't like what they were hearing from the CIA, so they created their own intelligence agency in Wolfowitz' office to come to preferred conclusions. I blogged about this here. The CIA did good work. As a unit they did not fall for the erroneous claims.