David R. Henderson  

Reynolds on Krugman and DeLong

PRINT
Caveat Emptor: Will The Buyer ... Introducing Alberto Mingardi...

Alan Reynolds has posted two excellent responses to criticisms of him by Paul Krugman and Brad DeLong. They are "Imaginary Squabbles Part 3: Krugman and DeLong's Changing Theories and Missing Facts" and "Imaginary Squabbles Part 4: Krugman and DeLong on the Top 1 Percent." My favorite of two is the second one above, probably because it addresses issues I have followed more closely.

You could easily spend hours clicking on all the links and then judging, against these links, both Reynolds' claims on the one hand and Krugman's and DeLong's claims on the other. I have spent only about 90 minutes and so it's possible that Reynolds made mistakes that I didn't notice.

Two key paragraphs from the second of the posts above:

The disgraceful thing about the October 15, 2011 CBO report was stopping the data with the cyclical peak of 2007 even though the CBO knew perfectly well that many "top 1 percent" businesses and investors were devastated by crashing real estate and stocks. CBO estimates that the top 1 percent's share of after tax income fell to 11.5 percent in 2009 - down from 17.4 percent in 2000 and unchanged from the 11.4 percent average of 1986-89.

Incomes of the top 1 percent did not bounce back after 2009, according to Piketty and Saez, merely rising from 18.1 percent of "market income" in 2009 (which arbitrarily excludes rising transfer payments from total incomes) to 19.8 percent in 2011. Piketty and Saez estimate that average real income of the top 1 percent was $1,322,635 in 2000 (measured in 2011 dollars) but real income dropped 20.7 percent by 2011 to $1,048, 234. Does a 20.7 percent drop over eleven years justify Krugman's anguish about "rising income" among the top 1 percent?


I would have to say, though, that Krugman's strangest claim is that people like Reynolds are "hired guns." Read Alan's response to this claim and judge the claim for yourself. I think it's strange not just in itself but in its origin: Paul Krugman is not exactly badly paid. He is hired. Is he hired to say the things he says? On an individual issue basis, no. But ask yourself this: if Krugman suddenly became a nice guy and starting arguing reasonably, granting points the other side makes, would his income go up, stay the same, or fall?


Comments and Sharing


CATEGORIES: Income Distribution



COMMENTS (25 to date)
Tom West writes:

His income would fall, strongly.

But that's because he would no longer be doing his job.

Just as a lawyers job is *not* to say, "you know, my client might be guilty", Paul Krugman's job is not to provide a fair hearing of the facts.

Belittling him because his job requires that he be partisan seems missing the point. As is expecting even-handed policy analysis.

David R. Henderson writes:

@Tom West,
I'm not belittling him. I'm pointing out the irony.

Tom West writes:

Sorry,

(1) I was actually referring to the general comments from us peanut gallery-types when Krugman is involved, not to your post - you're pretty much always quite generous in your posts, and

(2) Belittling is not the right word anyway - I meant criticizing.

Steve J writes:

I can understand trying to argue that inequality does not slow economic growth (even Krugman is hesitant to make that claim) but why argue that inequality does not exist? Isn't that the equivalent of arguing that The Great Stagnation did not happen?

David R. Henderson writes:

Steve J,
but why argue that inequality does not exist?
No one is arguing that. You did read Reynolds' post, didn't you? It was about the degree of inequality and the change in inequality.

Doug writes:

Reynolds is a hired gun but at least his position is relatively transparent about that fact.

Krugman works for Princeton and the New York Times. Two institutions that are far to the left of the median person. Two institutions that have consistently advocated left-wing positions for more than a century. Two institutions that wield an enormous amount of power and influence, certainly a lot more than Cato.

Yet unlike Cato, which even its supporters recognize as holding different views, the NYT purports itself to be "the paper of record." It will not even acknowledge its own progressive bias. Similar story holds for Princeton and every other major research university.

Hired guns for the NYT, Princeton and the rest of the progressive Cathedral are far more devious than their right-wing counterparts. At least right-wingers and their institutions are easily identifiable as such. The progressive Cathedral pretends like its only about unbiased truth-seeking. It holds the monopoly in our society on the ability to arbitrate mainstream scientific and credible views. In reality they have a far stronger hidden agenda than most others.

Of course when you actually confront them about these facts, they won't even make any apologies for it. But merely make condescending retorts like "reality has a liberal bias"

Steve J writes:

Sorry I was pulling a Krugman ;)

Actually I often agree with Krugman's ideas but admit he regularly ignores the subtleties of his opponents' arguments. So yes Reynolds is arguing the specifics of inequality rather than its general existence. But to some extent I consider this the same as arguing "the Earth is not warming as fast as they say" rather than "what are we going to do about this warming Earth". The second argument is much more interesting.

Steve J writes:

Doug,

You realize Krugman's blog is called Conscience of a Liberal right? Krugman does not hide his views at all.

And you must admit there is a much higher percentage of religious people on the conservative side. Reality and religion do not mix well.

David R. Henderson writes:

@Tom West,
you're pretty much always quite generous in your posts
Thanks, Tom.

David R. Henderson writes:

@Doug,
Reynolds is a hired gun
Please explain what you mean.

Jeff writes:

People with Paul Krugman's net worth complaining about inequality is quite something. "Chutzpah" is the word that springs to mind. What concrete steps do you suppose Krugman himself has taken to address the supposed problem, other than bemoaning it in a book that retails for $24.95? Charitable contributions, foregone tax deductions, unpaid labor, etc?

I imagine Paul Krugman makes substantially more than me. I think it's only fitting, in light of his bloviations on the subject, that he help alleviate the income inequality between the two of us by cutting me a check each month for half the difference between his salary and my salary.

Brian writes:

"Reality and religion do not mix well."

If that were so, religion would have vanished from humankind a long time ago. It's impossible for any institution to last thousands of years if it doesn't have a universal positive value added. Either religious belief is strongly aligned with reality, or reality doesn't really matter much. Since the latter seems highly improbable, the former must be true.

Ken B writes:
Reality and religion do not mix well."

If that were so, religion would have vanished from humankind a long time ago.


Non sequitur. Reality can vanish instead. And from at least some religious minds, it does.
Roger Powell writes:

(While fully admitting we've drifted off topic from Krugman vs. Reynolds, the last couple of comments moved me to make this recommendation.)

Regarding "Reality and Religion" - Some "religion" indeed survives and thrives while some variants don't. For an excellent talk about the "Economics of Religion", I recommend Russ Robert's 2006 interview with Larry Innaccone at EconTalk: Here

It's an informative discussion that thoughtful people of all persuasions (atheists, fundamentalists, etc.) will find stimulating.

Chris H writes:
If that were so, religion would have vanished from humankind a long time ago. It's impossible for any institution to last thousands of years if it doesn't have a universal positive value added. Either religious belief is strongly aligned with reality, or reality doesn't really matter much. Since the latter seems highly improbable, the former must be true.

I highly doubt the point you are arguing is correct. First off, I don't think a good case can be made that slavery created "universal positive value added" and yet it persisted for thousands of years. The reason it persisted is that while slavery creates large negative values to the large numbers of slaves, it creates significant positive value to the inherently smaller slave-owning class. That class maintains this net-negative equilibrium because it was capable of using violence better than the slaves were.

This is just one example of net negative equilibria. Another is the persistence of the Ptolemaic model of the solar system. This model persisted even though it was inferior to models that could have been conceived of given what was known about geometry and shapes had astronomers simply considered using elliptical orbits rather than circular orbits. Did humanity benefit from believing the sun orbited it more than vice-versa? That seems unlikely given that when the real nature of orbits was discovered and accepted there doesn't seem to have been any real disruption in the lives of most people. Plus knowing the truth did lead to cool things like properly flying space probes.

So no, it is not impossible for an institution to last thousands of years without giving "universal positive value." An argument that long-lasting institutions tend to give universal positive value is better, but also not certain given I don't think there's reason to think that negative equilibria are less likely than positive ones. For instance a majority of the human population right now lives in countries with bad institution that have created economies far inferior to countries with better institutions (mainly OECD countries). Getting stuck in a bad rut economically actually seems more likely than getting on a good one.

John B. writes:

Re Ptolemy's model.

Whether it's inferior or not depends on the criteria used to judge quality. Ptolemy's model answers the question "where will a given planet be in the sky on a given date" far faster than Copernicus' model and isn't less accurate. Copernicus' model is better if we are asking "what would the Solar System look like from outside".

Brian writes:

Chris H,

Just because we view slavery as being wrong doesn't mean it didn't provide net positive value in its time. Slavery as it existed over thousands of years tended to be the result of conquerers looting the spoils of the defeated nations. In the case of slavery, the looted spoils were human labor. It's likely that slavery of this form played a major role in the formation of civilizations and many of the extraordinary advances we celebrate from antiquity.

Can I prove that slavery provided net positive value? No. But I don't think you can prove your objection either.

Note that slavery as a widespread and legal institution did not end until the industrial revolution and free-market capitalism came along, and then it crashed fairly rapidly. Did humanity suddenly become more moral? No. The REALITY of what was possible economically changed, so that slavery no longer provided net positive value. The rapid dissolution of slavery under the new conditions provides strong evidence that slavery DID provide net positive value under the old conditions.

Now regarding the Ptolemaic system, setting aside that it wasn't an "institution," if you think it didn't provide net positive value, you really don't know much about the history of astronomy. Was the Ptolemaic system "true"? Not in the way we view it now. But it was still extremely valuable for charting star and planet positions and played a fundamental role in organizing time (through the calendar) and in navigation. It may not have been "true" but it was deeply aligned to the observable REALITY of celestial motion, which was the source of its value.

I will grant your point that bad equilibria are possible, but I have severe doubts that they can last for millenia. Please note that the existence of a better system does not imply that the worse system does not provide net positive value. The question regarding npv is not what happens if the institution is replaced by a new system, but what happens if the institution is simply eliminated.

Brian writes:

"Non sequitur. Reality can vanish instead."

Ken B.,

Reality never "vanishes." It can be denied, but that denial can't persist unless the "reality" involved doesn't matter (which was the second option I mentioned). In the case of creationism/ID (in your link), it did not persist among scientists because that reality mattered to them. It DOES persist among non-scientists because the reality of human origins really has no impact on the daily lives of people.

So my argument is not a non sequitor. It follows logically as given as long as we are focused on reality that matters.

Aaron Zierman writes:

Krugman is politically astute. He often makes the classic move of blaming the competition of what he himself is doing. Such as being a "hired gun" and grandstanding on issues while ignoring true debate.

Claiming that the debate is over and that your position is the correct one is no substitute for a real argument.

If Paul Krugman were "a nice guy, and starting arguing reasonably, granting points the other side makes", well, then he would no longer be Paul Krugman. Since he gets paid for who he is, and who he is gets paid relatively highly, it stands to reason that the person he would be in such a case would be paid less.

Ken B writes:

Brian,
If I suggested you rent a sense of humor would you object that it's impossible, really?
You asserted that if "religion" and "reality" really did conflict that religion would "vanish". That's just not so. Reality and astrology conflict and astrology hasn't vanished. I made a play on words to emphasize why you are wrong: people can have "reality" banished from their thoughts with religion controlling their minds.
It's pretty common actually. God is a virus that infects brains, crippling some of their functions.

Ken B writes:

[Comment removed pending confirmation of email address and for rudeness. Email the webmaster@econlib.org to request restoring your comment privileges. --Econlib Ed.]

JMK writes:

"if Krugman suddenly became a nice guy and starting arguing reasonably, granting points the other side makes, would his income go up, stay the same, or fall?"

No change. Sticky wages.

Chris H writes:

@Bryan

Just because we view slavery as being wrong doesn't mean it didn't provide net positive value in its time. Slavery as it existed over thousands of years tended to be the result of conquerers looting the spoils of the defeated nations. In the case of slavery, the looted spoils were human labor. It's likely that slavery of this form played a major role in the formation of civilizations and many of the extraordinary advances we celebrate from antiquity.

Slavery however clearly wasn't required for many of those. If Egypt could build the pyramids without slave labor (at most using temporary conscripted labor which you might say is still slavery, but it was of a very temporary sort not the life or nearly life types of slavery which existed) then it seems likely to me that slavery wasn't needed for society. And even beyond that, is value of having more nice looking ruins worth the enormous human tool it cost? And if we hadn't had slavery to misdirect labor from people trying to find ways to improve everyday life towards grandiose projects or armies to kill one another or service to small rent-seeking elites could have produced a better world than we currently have.

You say you doubt I can "prove" that slavery produced net negatives. But when a system produces large costs on a larger group and perhaps large gains on a smaller group the general trend should be that the costs outweigh the benefits. The burden of proof would lie with the person that the benefits to that small group are so large as to outweigh the costs to the large one. And we know of societies that had little to know slavery at about the same technological level as the ancient world and we know they often did rather well. Rome and Han China both had some slaves but for the Han slavery was only a small portion of the populace while slaves in Italy made up between 35-40% of the population. Both societies were comparable in wealth and size, but not having a large populace of people for whom there was little to no ethical consideration given should mean the people living under the Han had a better chance of living a more enjoyable life than people living under Rome. Sadly we don't have the data to absolutely confirm this, but again burden of proof should lie with the person saying oppressing large numbers of people produces a net benefit.

But of course, we do need to talk about net benefit compared to what. Which also dove tails well into a discussion of the ptolemaic model:

Now regarding the Ptolemaic system, setting aside that it wasn't an "institution," if you think it didn't provide net positive value, you really don't know much about the history of astronomy. Was the Ptolemaic system "true"? Not in the way we view it now. But it was still extremely valuable for charting star and planet positions and played a fundamental role in organizing time (through the calendar) and in navigation. It may not have been "true" but it was deeply aligned to the observable REALITY of celestial motion, which was the source of its value.

The Ptolemaic model was superior to any previous model that had existed yes. I'm not disagreeing with that. But the argument you made is that it's impossible to have an institution (and arguably if religion is an institution then the ptolemaic model was as well with schools dedicated to propagating it and even hierarchies of people working on it, though said hierarchy was primarily limited to teachers and students) which lats for millenia without producing a net benefit. But net benefit compared to what? Compared to a hypothetical situation of the worst case scenario? That's not a particularly strong argument in favor of long-lasting institutions. If that's the case then we should actually assume that most long lasting institutions are not optimized to produce the most benefit possible for a given situation/time. This is because in almost all situations there are a lot more ways to fail to be optimal but not optimally bad than ways to be optimally good (exceptions to this would be binary or trinary choices but binary/trinary choices are a lot more common on school tests than real life). I don't think that's what you're arguing however.

Perhaps you mean net benefit compared to some average option. Institutions produce net benefit compared to the average benefit other options would do so. Again this leaves plenty of room for improvement and we should still suspect institutions that last long periods are not optimal in a given situation (and thus religion is likely not an optimal institution given the knowledge and situation we find ourselves possessing and thus is likely not optimally approaching truth given what we know), though better than most options. This type of view may be correct (simply because there are so many ways to fail) but adherence to this too strongly can prevent people from coming up with better ideas.

The Ptolemaic model worked better than anything else proposed at the time, but that was because people were wedded to another long-standing and completely wrong idea, namely that the orbits of planets were circular. To be fair, this is close to true, but too much devotion to that idea prevented anyone from experimenting with other potential orbital configurations. Ellipses weren't unknown in the ancient world and if some astronomer had tried using those for orbital paths, Aristarchus' model would have proven superior to any combination of crystalline spheres Ptolemy could come up with and astronomers could have focused their efforts on answering other questions than "how many spheres do I need to add to make these predictions work" and instead asked questions like "wait, if we're circling the sun then just how far away must those stars be to not show any sign of parallax?!" That could have advanced astronomy and would have been a clear net benefit over a Ptolemaic model even without telescopes. Institutions fail to optimize for a given amount of knowledge very often which means that institutional longevity is at best a weak argument in favor of it's usefulness and even more so it's truth value.

Justin writes:

[Comment removed pending confirmation of email address and for rudeness. Email the webmaster@econlib.org to request restoring your comment privileges. Flame wars are prohibited on EconLog.--Econlib Ed.]

TMC writes:

Stephen and Ken B,
I'd argue that if you'd replace 'religion' with 'progressivism' you would be much closer to 'reality'

Comments for this entry have been closed
Return to top