21
Jan
2015
Correcting Hartmann on U.S. Tax History
My latest at Mises CA. On Facebook I saw a meme with 60,000 shares from “U.S. Uncut” that was simply awful. You don’t realize how awful it is until I walk you through it. Come, let us explore together.
Wow, that was awful. On the bright side, I now know if there’s a recession in the next several years, I can just blame it on Obama’s tax increases.
SOOOO increasing tax rates causes the economy to expand. Who knew? Apparently the solution of our current economic malaise is to confiscate 100% of our labor. Sunshine and rainbows to follow!
Perhaps Gruber was right.
Bob, didn’t the U.S. economy bottom out in mid-1932, just when the Hoover tax increases were signed into law?
No, not according to employment or industrial production figures.
E.g. Scott Sumner can make a case that FDR going off gold prompted the rebound in 1933, while progressives typically talk about his spending increases.
Look more closely at the Industrial Production figures, Bob. They do show the economy bottoming out in July 1932, but only having a truly sustained recovery beginning in April 1933. David Stockman undoubtedly exaggerates the “Hoover recovery”, but there is a grain of truth to his claims about it.
http://research.stlouisfed.org/fred2/series/INDPRO/
Spending increases in the first five months of the Roosevelt administration were not nearly large enough to explain the strength of the recovery at the time. The most important factor to the recovery probably was the restriction of gold exports from the U.S. in April 1933.
Yikes! I’ll do a whole post on this E. Harding, it serves to underscore my “don’t trust the experts” post. I was just testing you?
Also, there were some pretty bad boom-bust cycles back before 1913
http://research.stlouisfed.org/fred2/series/A0107AUSA322NNBR
But we know they’re actually worse now, thanks to work from the 90’s by Christina Romer(!)
Selgin White and Lastrapes’ paper on the Fed breaks down pre-1913 and post 1913 (even post 1945) comparisons. From what I’ve been able to gather, there was scarcely a better period for economic stability and growth in US history than ~1879-1913. Otherwise known as the “Classical Gold Standard.”
On the bright side, there was no worse period than 1913-1945, so we’re doing better now than then, at least.
Aw, come on, LK refuted that on his blog just a few days ago:
http://socialdemocracy21stcentury.blogspot.com/2015/01/libertarian-gold-standard-myths-never.html
The Reagan era had stronger growth. I also have the impression the National Banking System era was very unstable (e.g., 1907, 1893, 1896, 1873), but I might be wrong.
The National Banking System did cause many problems, and there were several financial crises/panics. But these don’t seem to have lead to severe prolonged depressions, at least in the Gold Standard (that is, international Gold Standard) period (ie post 1979). Even the characterization of the period from 1873-1879 as a “depression” is extremely dubious.
Note that there were no actual national surveys on Unemployment during those times, so estimates of unemployment depend on the statistical estimate of the “Output Gap.” Which requires one to estimate both GDP (or GNP) and the “trend.” I actually did something like the latter with Samuel Williamson’s data along with data on government spending from Christopher Chantrill, and some statistics (because the state/local level spending had to be estimated before about 1890) to estimate the “gap” in “private output” (actually GDP-Government Spending, which also removes transfer payments, but I don’t think that would change much about the result) and got this basically. You’re not going to be able to get high unemployment numbers from that (and this data correlates well with unemployment except during the Second World War, for periods we actually have data) for any year from 1879-1913. Even the worst years you’d probably get about 6%.
Also I misspoke, Romer’s work on Unemployment specifically was from the 80’s, although later work addressed output directly and there is also this more recently:
Romer, Christina D. 2009. ―New Estimates of Prewar Gross National Product and Unemployment. Journal of Economic History 46(2) (June): 341-52.
And if you look at the paper I mentioned earlier for more “mainstream” estimates of unemployment (figure 6), the cite for most of the period is:
Vernon, J.R. 1994a. ―Unemployment Rates in Postbellum America, 1869-1899. Journal of Macroeconomics 16(4) (Autumn): 701-14
Which is based on older GNP estimates, mind you! Not Romer’s GNP series that is less volatile. You’ll see though that that series never goes above 10% unemployment, something that has only happened in the more contemporary period, where we have A. A Central Bank and B. No International Gold Standard of the likes of 1879-1913. My own estimates would have that happening only twice from 1789 to 1913, both before the classical Gold Standard (and one is clearly associated with the Civil War). But those are my estimates, mind you. I think my treatment of the data is defensible, but there are mainstream cites which reach similar (if less striking) conclusions.
I should also add (assuming Bob fishes out my other comment) that describing “periods” I had in mind periods longer than one or even two decades. This is what is really necessary to properly characterize a “monetary regime” as opposed to a more fleeting political regime. The “interwar period” is 32 years, the International Classical Gold Standard is 34. The Reagan Administration is just 8.
Then you’re probably right on growth, but only if you don’t adjust for population or labor force growth. The late 1940s-early 1973 era is pretty much a tie, though. As for volatility, I think the Fed figured out the art of delaying recessions either in the 1960s or early-mid 1980s (Bernanke seems to have forgotten much of this in 2006-8, but re-learned in the aftermath of the Great Recession). Due to the lack of the FDIC, bank runs were much more common in the pre-1933 than post-1933 era. In terms of unemployment, however, I don’t think it’s implausible medium-term volatility was essentially unchanged from the early 1870s to today (though volatility on a monthly scale may have been greater).
It’s fair to point out, the growth rates I was talking about weren’t per capita. The highest 35 growth over a 35 year period of Real Per Capita GDP (Again, the measuringworth data) was the period ending in 1967 (starting in 1933) although, that’s partly so high because the beginning is at the trough of the Great Depression. So yes, in terms of growth rate of the productivity of the population, it’s better, broadly, after 1913 than before. I think to some extent that’s merely to be expected-improving, accelerating technology, for one thing. But to some degree the faster population growth owes something to the faster economic growth, too.
Either way, it’s one reason I wanted to look at business cycle variations as well as growth. Really long term growth should have relatively little to do with the monetary system. So the deviation of GDP from trend, and unemployment, are probably better indicators of the business cycle effects of the monetary regime.
Hard to say for sure about monthly (or even just sub annual) volatility. But it wouldn’t surprise me. The way the monetary system worked under the National Banking System was especially prone to causing seasonal problems.
Let’s see… anonymous internet troll runs a blog and purports to disprove sophisticated, peer-reviewed economic research with some spreadsheets.
Nah, I think I’ll stick with the professional work, not internet troll nonsense.
Makes me glad I cited sources!
I don’t think it would be wrong in principle for some guy on the internet to overturn academic work. I’ve tried to do that myself a few times! But (see Bob’s Feynman post!) always question claims of fact unless you can verify them for yourself.
Agreed, Andrew. I don’t have a problem with having a look at some stylized facts and making a general point about some broad concept or purely empirical question. However, time series econometrics is typically a significant improvement over that sort of thing.
LK seems to straw man his opponent by choosing a single metric and not asking his opponent what metric he used. The paper you cited (Selgin et al) discusses inflation, output volatility, bank failures, and other issues.
Although Hartmann wasn’t my target, the same nonsense was debunked in a recent video posted in my blog.
I looked at effective tax rates and focused a lot of attention on the Reagan years. Obviously the Progressives have an issue with their narrative when tax receipts grow despite all tax rates being decreased drastically.
This is what I mean when I refer to the “Conspiracy Theory of Economics,” people. Reasoning that goes something like: Rich People Buy Republicans -> Republicans Steal The People’s Money to give to Rich People (ignore the fact that it was the Rich People’s money in the first place) -> Rich People suddenly forget how to make money and crash the economy -> Rich People Profit somehow from not making money -> Rich People buy more Republicans.
You might notice this theory makes about as much sense as the world secretly being ruled by Jewish Reptile Men. That’s because it’s on about the same cognitive level.
A lot of economic “analysis” can be summed up by, “Booms are caused by [thing I like for other reasons] and recessions are caused by [thing I dislike for other reasons]. What a wonderful coincidence!”
1. When I encounter something like this I am often inclined to remind myself of the old saying about arguing with an idiot. Bob should always remind himself of that saying.
2. Having said that, Thom Hartman is obviously not an idiot. But when President Obama lies in the nation’s face and gets reelected for his trouble, why should anyone be worried about the truth? On a completely different story, Secretary Of State John Kerry recently said “France Invented Democracy”, and Bob wants to nitpick about some arcane tax history.
3. Anybody who believes the claim that changing one parameter in the economy, as important as tax rates are, and causing boom or bust within months is completely ignorant as well as ideologically set in concrete. That person won’t be changing his mind even if you bring them before a resuscitated FDR to certify the facts in person.
Still, for me this is a learning experience, being a non academic and surely not an economist. And I enjoy reading Bob’s columns. So may Krugman, Hartman and the rest keep provoking him.
Ah yeah if I was Greek I would have been pretty miffed. The French start cutting people’s heads off after the Americans develop a Representative Government modeled after but improved upon the already extant Representative Government in Britain, and over two thousand years after the Athenians had a Democratic Government, but yeah, John Kerry, France invented Democracy.
But I’m not Greek, so I just kind of laughed at what an idiot he is. Then I wept because there are probably people who think he’s correct.
There should be a televized debate between Hartmann and Piketty about who comes up with the best typos.
Bob, was my reply to E Harding eaten by moderation? I know I had a lot of links in there.
Probably not.
It’s there now, just looked for a second like it had disappeared from “waiting for moderation” status.
Andrew_FL:
But we know they’re actually worse now, thanks to work from the 90’s by Christina Romer(!)
No, Romer’s pre-1914 GDP calculations are NOT generally accepted, but rejected. It is Balke and Gordon’s estimates that are usually regarded as the best:
Balke, N. S., and R. J. Gordon, 1989. “The Estimation of Prewar Gross National Product: Methodology and New Evidence,” Journal of Political Economy 97.1: 38–92.
Levi Russell:
Let’s see… anonymous internet troll runs a blog and purports to disprove sophisticated, peer-reviewed economic research with some spreadsheets.
The data I present here:
http://socialdemocracy21stcentury.blogspot.com/2015/01/libertarian-gold-standard-myths-never.html
is taken straight from Angus Maddison’s The World Economy: Historical Statistics (OECD Publishing, Paris, 2003) — the standard specialist scholarly source for world GDP. . It is people like you are ignorant and who have no idea what you are talking about.
Romer’s work was on more than just GNP (But see my comment below) but:
I in fact also cited Vernon’s Unemployment estimates which are based on Balke and Gordon. Funny that, huh?
LK, my comment stands. Selgin et al is more thorough than your copy/paste of simple data plots. As Andrew notes, Selgin et al explains the alleged contradiction and does not conclude in your favor.
I’m not questioning your data source, LK, I’m questioning the robustness of your analysis. Andrew has done a good job documenting the recent research is not in your favor. Since Selgin et al explicitly lays out the problems with Balke and Gordon, you’ll have to deal with their analysis directly to make a convincing argument. You haven’t done so.
Andrew_FL
I don’t think it would be wrong in principle for some guy on the internet to overturn academic work.
There is no “overturning” of “academic work” by me in the post you cite.
The data is taken directly from the standard reference work on modern national GDP: . Angus Maddison’s The World Economy: Historical Statistics (OECD Publishing, Paris, 2003).
Maddison — just as many other economists — takes his GDP estimates for 1870 to 1914 from N. S. Balke and R. J. Gordon, 1989. “The Estimation of Prewar Gross National Product: Methodology and New Evidence,” Journal of Political Economy 97.1: 38–92.
Are you in the habit of appealing to authority rather than determining what the best data are yourself?
I’m going to assume so.
Andrew_FL
But we know they’re actually worse now, thanks to work from the 90’s by Christina Romer(!)
Selgin White and Lastrapes’ paper on the Fed breaks down pre-1913 and post 1913 (even post 1945) comparisons.
False.
If you had bothered to read Selgin et al.’s paper properly, you would have read:
“Romer’s revisions have themselves been challenged by others, however, including Zarnowitz (1992, pp. 77-79) and Balke and Gordon (1989). The last-named authors used direct measures of construction, transportation, and communication sector output during the pre-Fed era, along with improved consumer price estimates, to construct their own historic GNP series. According to this series, the standard deviation of real GNP from its H-P trend for 1869 to 1914 is 4.27%, which differs little from the standard–series value of 5.10%. Balke and Gordon’s findings thus appear to vindicate the traditional (pre-Romer) view … .” (Selgin et al. 2010: 11).
Balke and Gordon’s estimates are now widely considered the standard ones. It is you who have made a pretty shoddy error here.
Nice quote mining. Let’s dig a little further, shall we?
” More recent work helps to resolve the contradictory findings of Romer on one hand and Balke and Gordon on the other. Rather than rely on conventional aggregation procedures to construct historic (pre-1929) real GDP estimates, Ritschl, Sarferaz and Uebele (2008) employ ―dynamic factor analysis‖ to uncover a latent common factor capturing the co-movements in 53 time series that have been consistently reported since 1867. According to their benchmark model, which assumes that the coefficients (―factor loadings) relating individual series to the latent factor are constant, there was in fact ―no change in postwar volatility relative to the prewar [that is, pre-World War I] period (ibid., p. 7). Allowing instead for time-varying factor loadings (and hence for gradual structural change), Ritschl et al. find that post-WWII volatility was a third greater than pre-Fed volatility (ibid., p. 29, Table 1). These findings reinforce Romer‘s conclusions. But Ritschl et al. are also able to reproduce Balke and Gordon‘s postwar moderation using a common factor based on their non-agricultural real time series only, which resemble the series Balke and Gordon rely upon for their GNP estimates. Here again, the moderation vanishes if factor loadings are allowed to vary. Balke and Gordon‘s finding of a substantial reduction in post-WWII output volatility relative to pre-Fed volatility thus appears to depend on their focus on industrial output and implicit assumption that the relative importance of different components of that output hasn‘t changed.”
Gee, one wonders why, in holding up Balke and Gordon as the “Gold Standard” (pun intended) in historical GNP estimates, you didn’t mention the rest of this part of the paper just now? Maybe because it would make you look ridiculous.
.
Well said, Andrew.
You’re giving Hartmann far too much credit simply by acknowledging him. It’s like listening to a two year old: he has absolutely no concept of time or proportion: Harding caused a bubble that lasted nine years…Roosevelt presided over a booming economy (?)…Reagan caused a recession – what 2 years later? 70 years later? …Clinton caused a boom, Bush a bust, and we had 8 years of joy and prosperity because Obama raised the top rate from 36% to 39%?
If the cherry picking weren’t so blatant, we’d at least be able to show prof Krugman that austerian obviously works.
If the cherry picking weren’t so blatant, we’d at least be able to show prof Krugman that austerian obviously works.
Good one. But Krugman is also big on “marginal tax rates don’t affect economic growth.”
“I in fact also cited Vernon’s Unemployment estimates which are based on Balke and Gordon. Funny that, huh?”
And the method used by Vernon is highly questionable because we have strong evidence that labour force participation rates were countercyclical before 1914:
http://socialdemocracy21stcentury.blogspot.com/2013/12/us-unemployment-in-1890s-who-is-right.html
If true, those estimates by Vernon are underestimates, perhaps even gross underestimates, of the real unemployment rate.
“Selgin et al is more thorough than your copy/paste of simple data plots. As Andrew notes, Selgin et al explains the alleged contradiction and does not conclude in your favor.
Selgin et al does not settle the matter. In fact, a careful look at that paper shows they hedge their bets carefully. There is just as good a case for Balke and Gordon, and many feel a better case than Romer’s estimates.
That’s a somewhat bizarre reading of “Balke and Gordon‘s finding of a substantial reduction in post-WWII output volatility relative to pre-Fed volatility thus appears to depend on their focus on industrial output and implicit assumption that the relative importance of different components of that output hasn‘t changed.”
I think most people’s visceral reaction would be “That implicit assumption makes no sense.”