DISQUS

LarryM · 2 years ago

Will, you're a smart guy, and increasingly my ideological POV is moving somewhat in your direction, but this post exemplifies your weaknesses as a blogger. I'd summarize it as your penchant for biting sarcasm often triumphs over careful analysis, though that may be a bit oversimplified.

Anyway, I only have a couple of minutes before I have to pick up my daughter, so let me just hit some of the high points.

Now, if it had as much as 1 percent higher on average, that’s going to be close to the difference between GDP per capita tripling rather than doubling over that time span — which is to say, the difference between what we got and twice as much.

Just plain sloppy. One and one half times as much, not twice as much (3x/1.5 = 2x).

(Anyone know of a rigorous estimate?)

Hmm. I don't pretend to be an expert in such things, but you do. Yet not only don't you have a "rigorous estimate" handy, but you would rather wax sarcastic than look one up. I'd wager that it would be a heck of a lot lower than 1% per year.

U.S. GDP per capita is now about 30 percent higher than much of Europe’s. There is good evidence that this has to do with incentives to work, definitely including tax rates.

"Including" does a heck of a lot of work in that sentence. There is a considerable body of thought that by far the largest factor therein is unrelated to tax policy - labor market issues, etc. That may be right or that may be wrong, but there simply isn't a consensus among economists that tax rates make nearly the difference that you assert.

More generally, by picking at this one paragraph you miss a lot of important context. In the actual world, as opposed tot he world of Will Wilkinson's policy preferences, lowering taxes has meant increasing deficits, not lowering spending (yes, overs simplified, but basically accurate). This presents very real trade offs, including probably at least some countervailing negative effect on growth, and, more tot he point, the deferral of those taxes to a later date, where they will (assuming a negative effect from taxation) eat up some of those economic gains.

Also No one, least of all Chait, argues that we should go back to 70% marginal tax rates. Now one could argue that, whatever its excesses, the Republican tax cut movement has at least given us those lower rates, and that absent such movement we would be stuck at those rates. But that's quite a different argument than the one that you are making.

HispanicPundit · 2 years ago

Then there is the added fact that Chait seems to confuse effective tax rates with actual. Sure the effective tax rate might have been 91%, but since this was the time before Reagan closed all of them loop holes in the tax code the actual tax rate was something far lower.

LarryM · 2 years ago

Playing around with a compound interest calculator, I am pretty sure that your numbers are wrong even beyond your obvious math error. 1% more growth per year, compounded, yields roughly 35% more growth over a 30 year period - not 100%, not even 50%. Now certainly even that could not be fairly called "relatively minor." But let's run the numbers with some other assumptions. Let's say that tax cuts have increased growth by, say, .5% per year, partly counterbalanced by, say, a 0.2% decrease in growth due to increased deficits. (Yes, I pulled those numbers out of my ass, but then you pulled the 1% number out of your ass. I strongly suspect that even my estimates are too generous, that the real difference is smaller). That would yield, over 30 years, an economy just a tad more than 9% bigger.

Now I think we are in the realm of arguing that such difference is "relatively minor." Especially since even that can't be viewed in isolation. That additional growth has meant either an increased deficit (while I estimated the short term cost of that, I didn't estimate the long term cost, i.e., at some point we have to repay the principle) or cuts in spending. Now, someone like you is going to find those cuts on balance a good thing, and, in some instances I would even agree with you, but if we are doing a cost benefit analysis, those cuts certainly do come with a cost (and, to the extent that you disagree with this, that's an ideological, rather than an economic argument).

One last point. While some people on the left underestimate the effects of compounding on a growing economy, I'd also say that some people on the right underestimate the effect of compounding upon long term debt. I'm not a deficit hawk, I agree that the short term effects of deficit spending are probably not THAT high, and I even think that we could maintain a modest deficit almost indefinitely without too high a cost. But if you look at long term budget projections, compounding means that, absent very large increases in taxes, decreases in spending, or growth rates MUCH higher than the historical norm, we are looking at a situation down the road where debt service will be so large that we'll have the worst of both worlds - high taxes with most of the spending going to debt service.

Now I know the almost rote response - the entitlement crunch was coming anyway, even before the Bush tax cuts. True enough, but why anyone thinks that that is an argument in favor of said cuts is beyond me. The tax cuts just compounded the problem.

Unless, that is, one is wedded to the "starve the beast" theory of tax cuts. Which is a plausible (if, IMO, ultimately misguided) position, but, to the extent that that represents your real disagreement with Chait, hardly merits your sarcastic, poorly supported attacks.

LarryM · 2 years ago

Uggh. I need to be more careful with phrasing myself. for 35% more growth, read 35% larger economy.

GilM · 2 years ago

"Now, if it had as much as 1 percent higher on average, that’s going to be close to the difference between GDP per capita tripling rather than doubling over that time span — which is to say, the difference between what we got and twice as much."

Just plain sloppy. One and one half times as much, not twice as much (3x/1.5 = 2x).

I think Will meant that the growth would be twice as much. I initially thought it was an error as well. It could have been written more clearly.

But, this is a blog post, not a research paper.

LarryM · 2 years ago

Gil,

Point taken. I think that the REAL problem here isn't so much sloppiness or even Will's sarcastic manner. The problem is that Will is going outside of his area of expertise - and, even if it's just a blog, one should probably exercise a little more rigor than usual when doing so. Because frankly, at the risk of killing a fly with an anvil, this post and the previous one are not up to his usual standards.

Sigivald · 2 years ago

Larry: I think Will meant twice as much increase (3x-x = 2x, vs 2x-x = 1x, where x is the starting GDP), not twice as much GDP in the end state.

At very least, that was my reading of his statement, which can admittedly be parsed in either direction - but given that both are possible, I choose to err on the side of charity and assume Will meant the reading that's accurate.

GilM · 2 years ago

By the way, I just wanted to add that I thought that this was a fine post.

I like Will's sarcasm.

Regardless of the specific hard-to-come-by numbers, I think that the point that Chait seems biased to underestimate the effects of high tax rates is a perfectly valid one.

LarryM · 2 years ago

Tastes differ - I tend to think sarcasm is a valid argumentative tool only when the other side is making facially absurd claims. The particular claims made by Chait here, even if one disagrees with them, as one reasonably might, are not facially absurd.

I happen to think that its problematic even when my "side" uses sarcasm this way (though given the direction that my political views are going, my "side" is pretty darn small). There are plenty of conservatives and libertarians who write cogently on tax policy (see, e.g., Bruce Bartlett, even when he isn't criticizing the Bush administration) that I read with real interest. My take on Will in this case is that, right or wrong (and in the larger sense I do think that SOME of Chait's claims go too far, and there is certainly an argument that but for the Republican tax cut revolution, excesses and all, we would be in a worse place today in terms of tax policy), is that he is a smart guy with s superficial understanding of economics who has cherry picked those sources which are most congenial to his political and philosophical views.

Now, that's a pretty darn common phenomenon, and you certainly could accuse Chait of it as well. In a sense, the fact that I'm going on about it at such length is a sign of respect for Will. I expect better from him.

William Newman · 2 years ago

Gah, this seems to be a hard thing to write clearly about. I'm confused by others' attempts, and trying to explain clearly myself, I'm not doing so well. I miss LaTeX and a shared mathematical vocabulary.:-( But I will try.

I, like Larry and others, found the twice-as-much-growth thing confusing. I also reached the same conclusion as Sigivald independently, without reading the comments. I think both sides are right: it is confusing, but not such nonsense that one can't figure it out after "wha?"

I think the "close to the difference between GDP per capita tripling rather than doubling over that time span" looks correct. I don't know what compound interest calculator LarryM refers to; I will try to calculate it from first principles. It's a 36-year period. Raising 1.01 to the 36th power gives 1.43. One of the things that number means is that if the GDP of country X starts equal to country Y, then grows 1% proportionally faster per year for 36 years, afterwards the GDP of X will have grown to 1.43 times the GDP of Y; we could say it has grown to 43% larger. Now if country X is the United States of Imagination with some policy which lets it grow 1% faster, and country Y is the United States of real-world America what happens? USA1973 has grown to 2.00 times the GDP of USA1947; ISA1973 has grown to 2.00*1.43=2.86 times the GDP of USA1947.

If by some happy miracle my publicly posted calculation doesn't contain a stupid error, then "almost 3" seems like a fair paraphrase of 2.86; given that the answer depends on a rate of 1%-for-the-sake-of-argument, rounding to 1 digit of accuracy seems more reasonable than reporting "about 2.9" or "about 2.86".

Will Wilkinson · 2 years ago

Guys, Sorry for the imprecision. I meant the difference between two doublings versus three doublings in national income in a given period. The heuristic is you can tell how quickly GDP doubles at a particular growth rate by dividing that rate into 72. (See here: http://www.econlib.org/LIBRARY/Enc/EconomicGrow...). A 4% rate doubles the economy twice in 36 years. a 5% rate doubles three times in a bit more than 37. That's slightly longer than the range I was talking about in the post, but I meant to be making a general point about the size of the effect. Three doublings versus two leaves you with twice as much.

LarryM · 2 years ago

The difference in my calculation was caused by my focus on Will's "30 years" in his third paragraph. But, yes, previous paragraphs refer to a 36 year period, so there is some sloppiness there, on both my part and Will's.

But really this seems like a silly thing to focus on (yes, I refer to my own contribution as well, to a certain extent), given that the 1% number comes out of thin air.

My question - and it's a serious one - are there reputable estimates as to the real average yearly increase to GNP growth during those years would have been with different tax policy? Will asks a similar question, though he says "much lower tax rates," which is unhelpful for a couple of reasons. I'd like to have answers to two seperate questions: (1) how much additional growth would we have gotten with a more rational (but revenue neutral) tax structure? Rationale in the sense of lower marginal rates and fewer deductions (which is not NECESSARILY more regressive, particularly if you eliminate many deductions). (2) How much additional groth do we get from lowering taxes across the board.

The sense I have - and I'd love to see research to conirm or deny this - is that, tot he extent that we see significant benefits from changes in tax policy, the effect of a more rational tax structure is far greater than the effect of lowering the aggregate rate of taxation. Now, obviously the implications of this aren't WHOLLY comforting to "liberals," as it has at least some implications regarding progressivity, but there are ways to smooth out marginal rates somewhat without abandoning progressivity.

LarryM · 2 years ago

Ugh, more poor proofreading on my part, hopefully it's all intelligible. The last sentence in the third paragraph should have read "How much additional growth do we get from lowering taxes across the board." That's still not entirely clear, and obviously the relationship isn't linear throughout the range of possible tax rates. But let's say the range between 15 and 25 percent of GNP for federal taxation - holding the tax structure constant, and taking all factors (e.g., increased deficits) into account, how much additional growth do we get for each 1% decrease in overall tax rates? The sense that I get is the answer to that question is "very, very little."

LarryM · 2 years ago

Also, obvious there will also be diminishing returns from lower marginal tax rates, i.e., the ralationship isn't linear there either. As string as the case is for lowering marginal rates from the 60 to 70% range, I get the impression that going from 40 to 35, say, doesn't get you much. The flat tax has always seemed to me, to the extent that it's sold as boosting the economy, as snake oil. But what, if anything, does the research say?

LarryM · 2 years ago

One more small point/request, this time proofread, and then I'll shut up (at least until I something to respond to). My impression is that these debates tend to occur, on both sides, in an almost evidence free environment. My own take, specifically with regard to the Bush tax cuts, is that the critics have the better underlying economic arguments. But the lack of rigor in the public debate makes it frankly very difficult for me to make that conclusion with any degree of certainty.

Somewhere in my responses to one of these threads I said that I suspected that most support of the Bush cuts was based upon either political concerns or a starve the beast rationale. But probably that was at least a little unfair. One thing that hasn't been discussed in these threads was the capital gains tax. I suspect that, to the extent that people support the Bush tax cuts on purely economic grounds, the capital gains tax cut is a big part of the reason.

But it's hard to be sure, because most of the PUBLIC arguments in favor of the Bush taxes cuts (I'm not talking so much about the first one, given that, as I recall, the budget was still in surplus, though even then there were doubts that it would stay that way) were facially pretty weak. I'd love someone here to come right out and say something like this: "look, we realize that most of the Bush tax cuts had a pretty poor rationale from an economic perspective, but on the whole it was worth it to get a cut in the capital gains tax."

Or is that expecting too much honesty?