More Chait Action

by Will Wilkinson on September 14, 2007

Jonathan Chait replies to my criticisms. He basically seems to me to say:

(1) Like the author of The Party of Death, I am completely confused by why anyone would think this book is a partisan hatchet job.

(2) Low taxes are actually good. I just have a hard time saying that clearly.

(3) George W. Bush’s tax policy made the system less progressive, and ran up the deficit, which is a problem for me for reasons I won’t disclose.

(4) People who wanted to privatize social security are either stupid or lying when they tell you that they would have made it more progressive. I could tell you why, but I won’t.

(5) The state really does own everything, and cutting taxes for the rich really is upward distribution. I have arguments to that effect, but I’d rather not debate it. Instead, I’d rather discuss my book in a way that takes all my wildly contentious normative premises for granted, while still pretending to be a pragmatic empiricist.

Meanwhile, I’ve added a follow-up post on why Chait’s class war thesis really is incredible.

Viewing 15 Comments

    • ^
    • v
    Oh man, now you’re picking on the "empiricism" thing again. What did this dude ever do to you?

    It's just not fair - he is a journalist after all. You should cut him some slack.
    • ^
    • v
    Oh BTW, Will, if you have not seen King of Kong yet you need to. I couldn't stop thinking about your theories on relative status hierarchies while watching it. That's all.
    • ^
    • v
    A lot of(most?) wealthly families get their money these days from the government, thanks to "privatization" pushed through by the Republicans.

    That tends to blur the difference between the parties.

    How many millionaires would see their incomes fall if, say, medicare was scaled back?
    • ^
    • v
    "A lot of(most?) wealthly families get their money these days from the government, thanks to “privatization” pushed through by the Republicans."

    Perhaps Alphie can enlighten me and list some of the industries that have been privatized by the Republicans.
    • ^
    • v
    Chait wrote "Look, economists agree that, all things being equal, lower marginal rates help long-term growth, and higher deficits hurt long-term growth. There are disagreements about how much either helps or hurts. The effects either way tend to be pretty small. They're second-order effects."

    In fact, second order effects don't need to be small: this should be mathematically obvious to most people who understand what "second order effect" means, and it should be emphatically confirmed by anyone who's spent any time using perturbation theory or asymptotic analysis (like lots of ex-physics folk, including me).

    It is quite true that ordinarily second order effects do become small compared to first-order effects as a perturbation becomes sufficiently small. In the problem under consideration, I think almost everyone would agree that second-order effects would be small with an income tax of 10% or so, e.g. But higher-order effects (second order, third order, etc.) tend grow in importance more than linearly with the size of the perturbation. Are the >40% marginal tax rates that Chait endorses sufficiently small that their higher-order effects are necessarily proportionally small? As I understand it, Chait scoffs at Feldstein's efforts to answer this question directly, so let's try another tack, appealing to consensus on a historically distant tax. As far as I can see, the ill effects of tariffs should be second-and-higher order in pretty much the same sense as the ill effects of marginal income taxes under discussion. Then looking at http://en.wikipedia.org/wiki/Smoot-Hawley_Tarif... I find "The Smoot-Hawley Tariff Act 'imposed an effective tax rate of 60% on more than 3,200 products and materials imported into the U.S.,' quadrupling previous tariff rates." Is Chait prepared to argue that at 60%, the higher-order harmful effects of the Smoot-Hawley tariff act can reasonably be assumed to be small compared to their revenue-raising effect? If so he seems indeed, as suggested by The Economist, to be in a group so elite "that by late afternoon even his shadow falls outside the charmed circle." If not, why is it safe to assume that the second-order harmful effects of higher-than-40% income tax rates must be small compared to their first-order revenue-raising effects? Obviously responses to tariffs and responses to income taxes are different in detail, but it's not obvious to me why response to income taxes should necessarily be vastly more linear than response to tariffs.
    • ^
    • v
    Here ya go, Dirty:

    "That accumulation of suburban wealth, local economists said, is a side effect of the enormous flow of federal money into the region through contracts for defense and homeland security work in the five years since the Sept. 11, 2001, attacks"

    http://tinyurl.com/q7vay

    If you are getting rich off government pork, you want lower taxes but higher government spending.

    And that's just what they've gotten over the past six years.
    • ^
    • v
    Alphie baby, I adhere to the traditional definition of privatization meaning that government-owned entities are placed under private ownership and control. Now it appears that corporate welfare and/or rent seeking is defined as privatization. I'm not surprised. It's a continuation of the premise that holds that reduction of the rate of growth of a spending program from 40% to 20% is in fact a 50% "budget cut". The growth of the glorious state is indeed inexorable. To parapharase President Nixon, "we're all privatized now".
    • ^
    • v
    William Newman,

    While I must say that I prefer your analysis to our host's (in that you're trying to make an argument from real facts, rather than just engaging in snide attacks that are low on evidence), I find one significant flaw. Even ignoring the fairly significant difference between 40% and 60%, and the difference between marginal rates and overall rates, why compare two different types of tax (income vs. tariff) when we have more relevant data from the 40s, 50s, and early 60s on income taxes? While obviously it's POSSIBLE that growth would have been significantly higher than the (historically high) growth that we in fact encountered, it DOES seem to suggest - well, no, prove - that high growth and high marginal rates are compatible.

    Now, that doesn't make Chait RIGHT, but it does make his argument at least plausible, and requires of his critics stronger arguments than have been deployed against him.

    In terms of Will's original post, rather than waste my time in a detailed response, I'll simply say that if one looks at the Chait post to which Will is responding, the unfairness of Will's post should be fairly evident. Again, I'm not even saying that Chait is right, and Will is wrong, it's more that Will just doesn't even really TRY to be fair to the arguments. I will touch upon one example - number 2 above. No one disputes that all else being equal lower tax rates are beneficial. The disputes are: (1) how beneficial? and (2) since all else is seldom - rather, never - equal - how to we balance the costs and benefits of lower taxation? I don't think that Chait is unclear on this point. If he doesn't spend a whole lot of effort belaboring the obvious point that there are some benefits to lower taxes in isolation, it's because its an obvious point. One might fairly take him to task for not delving more deeply into the trade offs, but that isn't the argument that Will makes.

    My own take on all of this - as a formerly center left liberal who is moving in a left libertarian direction (though the libertarian component was always there), is that there is indeed a strong argument that taxes should be lower, for the simple reason that people should be able to keep the money they make (with some limitations - I'm not all of the way there). But the pure economic/growth arguments in favor of lower taxes tend to be empirically fairly weak - much weaker than one would think based upon mere intuition.

    Which isn't to say that all taxes - or all tax schemes - are created equally in terms of their effects on the economy. Obviously at some point the overall tax burden grows to the point where you DO see some pretty large effects, and some taxes (regardless of the overall rate as a percentage of the economy) are particularly pernicious (see, e.g., Smoot-Hawley).
    • ^
    • v
    The Dirty Mac,

    Point taken. The sad fact is that, in practice, privatization consists ENTIRELY of rent seeking and corporate welfare. I'm not MERELY being snide here; can you really think of many examples of privatization in the United States over the last which meet YOUR definition?

    I don't have an easy answer to this problem, other than the (fairly obvious, to me) answer that: (1) those tasks which must be done by the government (and wherever you fall on the political spectrum, virtually everyone agrees that there are some) should be done by the government, directly, whereas those functions that shouldn't be done by the government should be fully privatized.
    • ^
    • v
    Is alphie seriously arguing that defense contracts are creating "(most?) wealthy families"? In the suburbs?

    I mean, news flash, man. DC is not a microcosm of the nation as a whole. That the DC suburbs are the nations's wealthiest means nothing regarding the nation as a whole, or to the general source of the wealth of most wealthy Americans - note that the other locations in the list are Silicon Valley, Manhattan, and San Franciso, not exactly known for their Defense and DHS spending!

    (You also didn't quote "Basu, chairman of Sage Policy Group, an economic consulting firm in Baltimore, said the region's affluence was "decades in the making." During the 1990s, he said, suburban Maryland and Virginia emerged as hubs of information technology, biotechnology, and defense and aerospace work. And the high levels of spending since 2001 on defense and homeland security "have generated revenues for existing businesses and prompted business formation, job growth, income growth and investment."" either, oddly.

    I guess decades in the making, and all that IT and biotech stuff just doesn't support the thesis.

    Nor, as Mac said, was there any "privatization" to speak of, pushed through by those awful Republicans.

    The only thing vaguely close I can think of recently is the change to having military support operations being run by contractors rather than troops... and that dates back to when a Democrat was running the White House.

    Shall we re-cast the "peace dividend" reductions as a Republican plot?)
    • ^
    • v
    I do, though, have a question here: do any of Will's libertarian readers want to take a stab at defending the time of "privatization" which means services paid for and contracted by the government, but provided by private entities? I can't really see how that type of privatization can be defended upon libertarian gounds, as it seems to me that all of the normal inefficiencies attributed (often accurately) to government are just compounded by additional opportunities for rent seeking, without the compensation of meaningful market competition.

    Note that my above formulation would not include voucher systems (not "contracted by" the government. I certainly DO see the libertarian argument in favor of such systems, while also seeing counterarguments/practical problems.
    • ^
    • v
    LarryM: The first two paragraphs of your post are about things I wrote, as opposed to what Will Wilkinson wrote. There are three things in those paragraphs I'd like to respond to.

    First, "high growth and high marginal rates are compatible" and "that doesn’t make Chait RIGHT, but it does make his argument at least plausible." My point was just that one particular argument for his conclusions was technically inavlid. That point doesn't depend on my disproving the conclusion. It seems as though you might have misread me, to think I was trying to disprove the conclusion. (That would not be an outrageous misreading on your part. I'm sure it's clear that I do disagree with the conclusion. And in choosing an analogy to a policy with widely-acknowledged second-order effects, I chose one where those effects are widely acknowledged to have damaged growth.) Really, I was trying to point out that the argument is technically invalid, independent of the conclusion. If someone writes "like all other geometric figures, kittens before the age 10 days have no more than five corners, therefore there the set of prime numbers is infinite," the argument is invalid. Is there an infinite number of prime numbers? Yes, in fact there is, so the conclusion is correct, but the argument is still invalid.

    I don't think I'm just being a partisan in thinking this is important here. By coincidence, very recently a similar technical error was made in (for me) the opposite partisan direction: see the comments at http://volokh.com/archives/archive_2007_09_16-2... . Technically the two cases are quite parallel. (Here, Chait implicitly claims that the terms after the first in a Taylor series are negligible. There, Roth implicitly claims that the terms after the second in a Taylor series are negligible. In both cases, the implicit assumption is so extreme that it is unreasonable to make it without mentioning it and supplying some serious justification.) For me as a partisan, the cases run in opposite directions: I broadly oppose Chait's conclusions, and broadly support Roth's. (I think historians, and others like staff writers at the _Economist_, *should* have be embarrassed for having signed off on that bilge.) But the partisanness difference didn't stop me from seeing technically similar problems, and seeing them as similarly important. (However, on VC I didn't get to point out the problem. Here, I was fifth poster. At VC, that might happen on the day the moon loses her daughter, if that occurs in a week when two Mondays come together.)

    Second, once you set out to defend the logical validity of Chait's argument by defending the factual correctness of his conclusion, you appeal to high US growth in the 40s, 50s, and 60s to say "it DOES seem to suggest - well, no, prove - that high growth and high marginal rates are compatible. [...] Now, that doesn't make Chait RIGHT, but it does make his argument at least plausible [...]." That seems like damning with exceedingly faint praise. How many real-world economic policies are so stunningly bad that they pretty obviously screw up economic growth singlehandedly? Near-total state control of the economy is the single candidate I can think of. And even for a policy as all-encompassing and bad as that, I think many people wouldn't find the evidence convincing if it weren't for the crude controlled experiments in partition of Europe in general, Germany specifically, China, and Korea.

    It is entirely typical, when a policy which is well-understood to be inefficient, it can be said that high growth and SuchAndSuchHorriblePolicy are not incompatible. For example, racially discriminatory laws like Jim Crow and apartheid are not just morally objectionable, they have long been understood to be economically inefficient. (See http://www.econlib.org/LIBRARY/Columns/LevyPear... for some colorful history.) That inefficiency is not disproved by pointing out that various countries have grown rapidly despite them. Similarly, trade barriers are inefficient even though of course countries can grow despite them: there are there are lots of examples.

    In fact, for trade barriers, there are not just lots of factual examples, but the aforementioned Smoot-Hawley tariff makes an amusing counterfactual example. I've never heard a serious economist try to explain the Great Depression solely in terms of the Smoot-Hawley tariff. Thus, I'm probably in good company in this opinion: if other things had been right in the 1930s, the Smoot-Hawley tariff alone would not have been enough to kill economic growth. So it's as though this is the second time running that a defender of the efficiency of high income tax rates has chosen an argument which would suffice to defend the efficiency of Smoot-Hawley.:-|

    Third "that doesn’t make Chait RIGHT, but it does make his argument at least plausible, and requires of his critics stronger arguments than have been deployed against him." Really? Of course you are correct that it requires of his critics stronger arguments than claims that 60ish-percent income taxes always suffice to cause very slow economic growth. However, I don't think it is the case that not a single critic, anywhere, has put forward an argument which is stronger than that.
    • ^
    • v
    LarryM, re. privatization.

    The problems you refer to are quite real. Alas, I don't have a great reference on hand to show that that they are taken seriously by libertarians. Instead I will express my respect for one particular libertarian by betting that Alex Tabarrok took some trouble looking into them in his book about privatization of prisons. (haven't read it)

    Note, though, the difficulty of such problems seems to vary a lot between cases. Prisons seem like at least a moderately difficult case. In some cases, there can be even harder problems than you mention, like the technical difficulty of expressing property rights for many kinds of water supplies (huge aquifers, e.g.), or the technical difficulty in saying anything usefully precise beforehand about what you're buying (custom-written software for air traffic control, e.g.). On the other hand, there are various niches where those problems don't look very difficult. Privatizing mail delivery, for example, might be nearly as simple as just eliminating the laws which prevent services other than the USPS from doing ordinary mail delivery.

    Note also there is ambiguity in "privatization" as in "deregulation." Libertarians generally favor privatization and deregulation, but that doesn't mean that any damfool thing which its backers or critics call by those names is a good idea. Roughly, libertarians dislike prior restraint, central control, and rule of men as opposed to law. We prefer rule of law, freedom of contract, clear and tradeable property rights, and such. The electric power "deregulation" in California worked notoriously badly, and I have come to expect people to see that as a telling criticism of libertarian ideas. My usual response is that "deregulation" with arbitrary price controls set by some government board, isn't: it's like "democracy" where the elections are limited to candidates chosen by some government board. Similarly, when a libertarian wants to privatize the postal service, it probably means he wants allow anyone who wants to to offer postal services. However, what is often called "privatization" in working politics is often something quite different. If I hear a postal service is "privatized" in the next ten years or so, I will expect it to mean something different: there is still a state-imposed monopoly on mail, it's no longer given to a department of the state, but to some corporation or other.
    • ^
    • v
    Really? Of course you are correct that it requires of his critics stronger arguments than claims that 60ish-percent income taxes always suffice to cause very slow economic growth. However, I don’t think it is the case that not a single critic, anywhere, has put forward an argument which is stronger than that.

    Sure, some of Chait's critics made more sophisticated arguments than that. But I was responding to Will, and, to a much lesser extent, to you; my point really was that Will hadn't made a stronger argument than that, at least on that point. And, of course, your argument, while far more sophisticated than Will's wasn't even MEANT as a general refutation of his position, but, as you said, a critique of one particular argument.

    Moreover, while I'm sure they are out there, and maybe you can even point me to it, I haven't seen a really sophisticated defense of the Bush tax cut legacy AS A WHOLE, on economic grounds. When even Alan Greenspan criticizes it in retrospect - well, pardon me for suspecting that that's because such arguments are weak or non-existent. I suspect that the real reasons for support for the project as a whole fall into two categories: the crude political motivations (combined with the naive extreme supply side arguments that no serious, intelligent, person really believes and even Will condemns) that Chait suggests, and some version of the old "starve the beast" argument. I'm less dismissive of the latter rationale than many people are, but it would be nice to see some more forthright admissions that that is what's going on.

    On a slightly unrelated note (I'm speaking as a non-economist, but one who (a) was 3 credits short of an undergrad economics degree, (b) has self educated himself to become reasonably knowledgeable about economics, and, perhaps most importantly, (c) has a mathematical/statistical aptitude and knowledge base to be able to avoid some of the worst errors that people make in this area, the level of economic ignorance on both the left AND the right, as displayed in blog entries and blog comments, is on the whole rather depressing.
    • ^
    • v
    Regarding your other post, I don't doubt that those issues are taken seriously by libertarians, and in fact I'm familiar with many of those arguments. I suppose what bothers me, and pert of the reasons for my question, is that I have seen a number of examples where self described libertarians are all too ready to defend the worst type of privatization, along with relfexive support of any policy which superfically means "less governemnt," even when the policies run afoul of basic libertarian principles. Of course, many of those self decribed libertarians aren't, really, libertarians.

    (Parenthetically, this really exposes the problem with libertarian support for the Republican party - even on those issues where conservatives and libertarians are ostensibly on the same page, the resulting policy is often pretty rank from a libertarian perspective. Not that I'd expect the Dems to be much if at all better).
 

Trackbacks

(Trackback URL)

close Reblog this comment
Powered by Disqus · Learn more
blog comments powered by Disqus

View My Stats