As a purely practical matter, the lower rate will probably sunset in 10 years unless reuppes by congress, and that uncertainty will accelerate activity that would have occurred later to the period right before the rate ends. Happens all the time.
"How are you going to play Dota if your fingers and bitten off? You can't. That's how" -> Carnarvon
"You can be yodeling bear without spending a dime if you get lucky." -> reVerse
"In the grim darkness of the future, we will all be nurses catering to the whims of terrible old people." -> Hacksaw
"In fact, our whole society will be oriented around caring for one very decrepit, very old man on total life support." -> SKFM
I mean, the first time I met a non-white person was when this Vietnamese kid tried to break my legs but that was entirely fair because he was a centreback, not because he was a subhuman beast in some zoo ->yotes
Resurrecting this and paging @ronya@enc0re and whoever else need paging because apparently one of the two intellectual foundations of the modern austerity movement was discovered to be... well... bad at Excel, which mistakenly created a result that indicated that high debt to GDP ratio = lowered growth.
Link is to Krugman, Krugman links to the papers documenting the finding.
I read the paper earlier. Pretty damning. But let's give R-R a chance to respond first.
@ronya : Shouldn't this be done as a panel VAR to begin with? WTF is going on that they are averaging growth rates in Excel?
Ideally yes as far as I can tell. They probably would not need to do a full VAR if they made some assumptions but it probably would not hurt.
If you have panel data there isn't really much reason to not use it as a panel. In the case of this paper it looks as if they didn't do a regression on their data at all. And in addition they did write a conclusion as if their work had value. That is very suspect since you should not have conclusions if you haven't done the work yet. To me it looks like they did a data description and called it good. That being said I only see the working paper at home which may not contain more work which exists in the published version(was r/r published?).
Still looks pretty damning to me. If I were reviewing that piece as I see it I would not accept it for publication in its current state. Like seriously that is low undergraduate quality. They never even state their model.
I cannot believe that this is what the austerity folks got hung up on
Edit Krugman in his next post is discussing regression(i think) results so it may be the case that what we are seeng is missing the crucial price of the pie
it was a descriptive paper to motivate a choice of stylized fact. this is macro, you're allowed to have only-very-loosely motivated stylized facts.
in the book they went and did the the event studies and such, it seems. the paper itself was not intended to be rigorous. Apparently it was even less than not-rigorous, though.
Still looks pretty damning to me. If I were reviewing that piece as I see it I would not accept it for publication in its current state. Like seriously that is low undergraduate quality. They never even state their model.
I cannot believe that this is what the austerity folks got hung up on
Edit Krugman in his next post is discussing regression(i think) results so it may be the case that what we are seeng is missing the crucial price of the pie
that's not really fair, it's basically phillips-curve-type writing. "hey look, what a weird apparently-structural relationship that our book may explain. Now buy/cite our book. thanks."
a little scummy but it's a common tactic and the model is in the book.
it was a descriptive paper to motivate a choice of stylized fact. this is macro, you're allowed to have only-very-loosely motivated stylized facts.
in the book they went and did the the event studies and such, it seems. the paper itself was not intended to be rigorous. Apparently it was even less than not-rigorous, though.
Still looks pretty damning to me. If I were reviewing that piece as I see it I would not accept it for publication in its current state. Like seriously that is low undergraduate quality. They never even state their model.
I cannot believe that this is what the austerity folks got hung up on
Edit Krugman in his next post is discussing regression(i think) results so it may be the case that what we are seeng is missing the crucial price of the pie
that's not really fair, it's basically phillips-curve-type writing. "hey look, what a weird apparently-structural relationship that our book may explain. Now buy/cite our book. thanks."
a little scummy but it's a common tactic and the model is in the book.
None of that should really be kosher in a field that aspires to call itself science.
That's using your conclusion to find evidence, rather than the other way around. Kind of a big deal.
People have been trying to replicate the numbers for a while now and couldn't.
I mean, who would think that to do it you need to excise certain cells in a spreadsheet?
Fair point.
I just can't believe how blatantly unethical this has been. I mean, compare this to the "No Global Warming, We Promise" guys that were shouted down by the scientific community repeatedly and continually. It takes 5 minutes for an average person to hear both sides of that argument, do a bit of googling, and find out who's been paid off.
the R/R paper is bullshit but most of the layman responses don't get it either, which is annoying
oh well
They're acting/sounding like political operatives.
no, not really. I think they just really wanted to believe it, and the derivation is so mindbogglingly simple that almost nobody bothered to doubt whether it was real. the flaw was discovered when people finally got fed up at not being able to replicate it, and R/R cheerfully volunteered all the data and the original spreadsheets used to construct the argument. this doesn't suggest any deliberate manipulation
the thing is, in macro it is normal to observe patterns like this and then make strong statements. this is because (1) there is a strong prior in favour of the absence of patterns, particularly across countries, so any apparently strong pattern at all is surprising (2) moving from strong correlation empirics to weak causative theory is normal since a certain J. M. Keynes legitimized the practice. the term of art is "stylized fact". this is normal practice.
the idea is that it's actually pretty darn difficult to construct plausible stories that line up with all the stylized facts we know of - certainly harder than laymen think it is - so identifying good stylizations does half the work. the inflation-unemployment tradeoff got to drive policy for five decades including today, for instance.
the "ha ha, what a bad way to back up macro policy" thing is grating because almost all macro policy is like that, including the policies you like! where does everyone think favourable multiplier estimates come from?
+1
Options
SerukoFerocious Kittenof The Farthest NorthRegistered Userregular
People have been trying to replicate the numbers for a while now and couldn't.
I mean, who would think that to do it you need to excise certain cells in a spreadsheet?
Fair point.
I just can't believe how blatantly unethical this has been. I mean, compare this to the "No Global Warming, We Promise" guys that were shouted down by the scientific community repeatedly and continually. It takes 5 minutes for an average person to hear both sides of that argument, do a bit of googling, and find out who's been paid off.
"How are you going to play Dota if your fingers and bitten off? You can't. That's how" -> Carnarvon
"You can be yodeling bear without spending a dime if you get lucky." -> reVerse
"In the grim darkness of the future, we will all be nurses catering to the whims of terrible old people." -> Hacksaw
"In fact, our whole society will be oriented around caring for one very decrepit, very old man on total life support." -> SKFM
I mean, the first time I met a non-white person was when this Vietnamese kid tried to break my legs but that was entirely fair because he was a centreback, not because he was a subhuman beast in some zoo ->yotes
it was a descriptive paper to motivate a choice of stylized fact. this is macro, you're allowed to have only-very-loosely motivated stylized facts.
That is fair, but i was reading it outside having read their book. The methodology in the paper doesn't identify the stylized fact it just describes the data. I.E. it has motivation to run the run the data with a particular model but doesn't describe the model they're motivated to use or the results of the test.
I mean i can look at a data set and say "wow it looks like there is a regime change right here" but unless i test that i haven't identified a regime change.
it was a descriptive paper to motivate a choice of stylized fact. this is macro, you're allowed to have only-very-loosely motivated stylized facts.
That is fair, but i was reading it outside having read their book. The methodology in the paper doesn't identify the stylized fact it just describes the data. I.E. it has motivation to run the run the data with a particular model but doesn't describe the model they're motivated to use or the results of the test.
I mean i can look at a data set and say "wow it looks like there is a regime change right here" but unless i test that i haven't identified a regime change.
the problem is that R/R don't have a model that gives any reason to believe that there is a regime change that applies for all countries and all years in vastly different circumstances
what they had, and have, is a model that gives regime changes at different cutoffs for some regions and others. their earlier non-bullshit work basically made the well-duh argument that countries with high risk premia due to serial default tended to continue to have high risk premia for loooong times. so you can run regressions to estimate exactly what risk premia level they got stuck at. well, sure, why not, that's hardly surprising. but the model stops there: it just says that countries will have this cutoff that they can't really shift, it doesn't say what the cutoff is
but the later paper is every academic's fantasy: that the approach they discovered is not only generalizable, but generates this astonishing and dramatic result for a whole swathe of other data. it's really super convenient:
the status quo before R/R was that nobody knew what level of debt burden is sustainable or tolerable, clearly the naive OLG optimal-debt-level models aren't plausible given the tendency for stuff to explode for no apparent reason. So it's actually pretty dang amazing! Not so amazing that it is unbelievable - everyone concedes that there is some level of D/Y where it becomes incredible that anyone who buys in will be repaid. maybe advanced industrial economies all do wind up with the same-ish cutoff for structural reasons, who knows? That would certainly be a highly elegant outcome.
it's not: countries will have this cutoff wrt debt load. It's just: countries will have a cutoff, and gosh it does seem to be the same-ish cutoff, amazing.
like I posted earlier, this is "look at the Phillips curve! Look at it!" level of argument. You can't say that five decades of looking very carefully at Phillips curves hasn't been highly impactful; nobody needs to contend that all countries face the same tradeoff, or even that the tradeoff is structural, to claim that it is relevant to policymaking
But the Philips curve existed. It was the result of actual tests that have evidence that the result was not random. Certainly it appears they did that in their book but they did not do so in their paper. We are beyond "look at this Phillips curve" we have at least gotten to "it look like the relationship the Phillips curve purports to exist exists and is t just random"
At the most base level they should know that if they're trying to find the debt level cutoff for nations and theor theory thinks they should all be relatively different that averaging cannot give any evidence that a cutoff is universal. They would have to have been looking at the individual data for all countries in order to make that claim. Otherwise they have to say "this is the level for everyone but the UK" or whatever
But it is normal, particularly amongst less-rigorous policy discussions, to presume that (1) advanced industrial economies are structurally similar (2) the D/Y limit is structural in some way. Even well before R/R, people just used D/Y=1 as the Schelling-point of panic. Hell, Maastricht set D/Y=0.6, which is even more onerous than R/R's 0.9.
why is there/does there have to be a limit? isn't it basically arbitrary, for countries with their own currency?
also, how do you tell that you have gone over an acceptable amount of debt? can't the things (I guess bond prices would plummet, or something) that you might ascribe to having unacceptable debt levels also be blamed on other things?
why is there/does there have to be a limit? isn't it basically arbitrary, for countries with their own currency?
also, how do you tell that you have gone over an acceptable amount of debt? can't the things (I guess bond prices would plummet, or something) that you might ascribe to having unacceptable debt levels also be blamed on other things?
when you have any non-trivial amount of debt, you have to have to be able to roll it over: to issue new bonds to fund paying off earlier bonds. this is fine because governments are not households with a finite expected lifespan.
but rolling over the bonds requires investors who believe that they, too, will be repaid in inflation-adjusted terms when their time comes. if you can't convince them to keep buying, then you've run up too much debt, because now you owe far more money than you can plausibly raise in emergency taxes and levies. having your own currency doesn't help either, because whilst you can satisfy the nominal values of your debts, everyone realizes that you failed to repay in real terms, and treats you the same as if you had just defaulted
why is there/does there have to be a limit? isn't it basically arbitrary, for countries with their own currency?
also, how do you tell that you have gone over an acceptable amount of debt? can't the things (I guess bond prices would plummet, or something) that you might ascribe to having unacceptable debt levels also be blamed on other things?
when you have any non-trivial amount of debt, you have to have to be able to roll it over: to issue new bonds to fund paying off earlier bonds. this is fine because governments are not households with a finite expected lifespan.
but rolling over the bonds requires investors who believe that they, too, will be repaid in inflation-adjusted terms when their time comes. if you can't convince them to keep buying, then you've run up too much debt, because now you owe far more money than you can plausibly raise in emergency taxes and levies. having your own currency doesn't help either, because whilst you can satisfy the nominal values of your debts, everyone realizes that you failed to repay in real terms, and treats you the same as if you had just defaulted
What I meant by a limit being arbitrary is, as you say, the investors who previously were buying debt for a given country would have to stop buying that debt in order for negative consequences to manifest. They have to have a reason for deciding to stop buying debt. What constitutes a reason, for a western economy that is basically stable/isn't facing any special circumstances that might add risk?
why is there/does there have to be a limit? isn't it basically arbitrary, for countries with their own currency?
also, how do you tell that you have gone over an acceptable amount of debt? can't the things (I guess bond prices would plummet, or something) that you might ascribe to having unacceptable debt levels also be blamed on other things?
when you have any non-trivial amount of debt, you have to have to be able to roll it over: to issue new bonds to fund paying off earlier bonds. this is fine because governments are not households with a finite expected lifespan.
but rolling over the bonds requires investors who believe that they, too, will be repaid in inflation-adjusted terms when their time comes. if you can't convince them to keep buying, then you've run up too much debt, because now you owe far more money than you can plausibly raise in emergency taxes and levies. having your own currency doesn't help either, because whilst you can satisfy the nominal values of your debts, everyone realizes that you failed to repay in real terms, and treats you the same as if you had just defaulted
What I meant by a limit being arbitrary is, as you say, the investors who previously were buying debt for a given country would have to stop buying that debt in order for negative consequences to manifest. They have to have a reason for deciding to stop buying debt. What constitutes a reason, for a western economy that is basically stable/isn't facing any special circumstances that might add risk?
for the relevant levels of debt we are concerned about, mainstream macro generally argues that panic or lack thereof tends to be self-reinforcing. Sovereign debt default has a bank-run-esque dynamic here: if a lot of investors pull out, then the state will have trouble rolling over the debt, and so it makes sense for even more investors to pull out. This is argued to be why solvency crises tend to be sudden.
It can be nothing more than a panic: countries can, and have, simply suppressed it through capital controls (to prevent frightened investors from moving their funds out of the country, so that they have nothing else to buy with the currency) and IMF emergency loans (i.e., resorting to the lender of last resort) and then a few years on, bond repayments continue successfully, and it becomes apparent that there was no fundamental insolvency. The controls might then be lifted and the IMF loans repaid.
given such a dynamic, the reason can be pretty much anything.
0
Options
L Ron HowardThe duckMinnesotaRegistered Userregular
I really kinda wish I knew what you guys were talking about.
Where does one start to learn some of this stuff?
I really kinda wish I knew what you guys were talking about.
Where does one start to learn some of this stuff?
I teach college economics for a living. To the interested individual, I always recommend the following book. Macroeconomic Essentials. It is clear, short, and doesn't require a big math background. It also helps that it isn't very expensive. At the same time it manages to be rigorous. It's one of those books that you can understand the first time through, yet every time you read it you discover another insight that you didn't appreciate the first time.
Now the specific thing we were just discussing is closer to a topic called 'Econometrics.' This is the intersection of economic models and how you implement them on statistical data. There is an excellent book I can recommend on this subject that has gotten many an economist through their Ph.D. program, yet can be read by non-economists. A Guide to Econometrics.
You will notice that it is written by the same author. That's no accident. Peter Kennedy is the kind of economist who understands his field so well, he can explain even difficult concepts with crystal clarity. I read this Econometrics text in grad school. Because of how good it was I looked around what other textbooks he had written and found his macro book that way. I even taught out of it for a while. But I have found that it's better for studying on your own. For a course, students do better with a more comprehensive book that has problem sets, supporting websites, online homework problems, and all that jazz. I use N. Gregory Mankiw these days, same as just about everyone else. You know, this guy:
for the relevant levels of debt we are concerned about, mainstream macro generally argues that panic or lack thereof tends to be self-reinforcing. Sovereign debt default has a bank-run-esque dynamic here: if a lot of investors pull out, then the state will have trouble rolling over the debt, and so it makes sense for even more investors to pull out. This is argued to be why solvency crises tend to be sudden.
It can be nothing more than a panic: countries can, and have, simply suppressed it through capital controls (to prevent frightened investors from moving their funds out of the country, so that they have nothing else to buy with the currency) and IMF emergency loans (i.e., resorting to the lender of last resort) and then a few years on, bond repayments continue successfully, and it becomes apparent that there was no fundamental insolvency. The controls might then be lifted and the IMF loans repaid.
given such a dynamic, the reason can be pretty much anything.
I guess what I find interesting about all this furor over the validity of the 90% debt limit is, since basically anything can cause investors to panic, the appearance of some kind of debt limit could act as an actual debt limit, if the reporting and dissemination of the limit is strong enough. I think investors are probably more scared of the idea of abandoning american debt than doing anything about exceeding 90% in the case of the states, but if it had been relevant to another country, could that have been enough to cause the kind of panic you were describing? It just seems a little surreal.
I'd be inclined to give them an ethics pass on the spreadsheet error. But what looks more like cherry picking seems much more deliberate and unethical.
for the relevant levels of debt we are concerned about, mainstream macro generally argues that panic or lack thereof tends to be self-reinforcing. Sovereign debt default has a bank-run-esque dynamic here: if a lot of investors pull out, then the state will have trouble rolling over the debt, and so it makes sense for even more investors to pull out. This is argued to be why solvency crises tend to be sudden.
It can be nothing more than a panic: countries can, and have, simply suppressed it through capital controls (to prevent frightened investors from moving their funds out of the country, so that they have nothing else to buy with the currency) and IMF emergency loans (i.e., resorting to the lender of last resort) and then a few years on, bond repayments continue successfully, and it becomes apparent that there was no fundamental insolvency. The controls might then be lifted and the IMF loans repaid.
given such a dynamic, the reason can be pretty much anything.
I guess what I find interesting about all this furor over the validity of the 90% debt limit is, since basically anything can cause investors to panic, the appearance of some kind of debt limit could act as an actual debt limit, if the reporting and dissemination of the limit is strong enough. I think investors are probably more scared of the idea of abandoning american debt than doing anything about exceeding 90% in the case of the states, but if it had been relevant to another country, could that have been enough to cause the kind of panic you were describing? It just seems a little surreal.
I think so.
Back during the 1997 Asian financial crisis there was actually a vaguely similar theory floating about - some econometricians (Alywn Young, etc.) had studied the contributors of economic growth in the Asian tigers and found that much of the growth had appeared to come from adding more and more labour and capital (pulling women from households to factories etc.) rather than increasing the productivity of each added unit of labour and capital. The argument was thus that the growth would have to abruptly slow down once they ran out of women to shift to factories, and so the high predicted growth rates weren't really justified and that a huge banking crisis was going to occur once new investments turn out to be less profitable than expected. See, e.g., The Myth of Asia's Miracle, published in Foreign Affairs in 1994 by one of the world's most influential international trade economists, a certain Paul Krugman (yeah :rotate: . Note that Krugman himself correctly diagnosed the crisis when it then hit, although many of the theory's supporters did not). The intimation was that, like the Soviet Union, the pace of growth was straightforwardly unsustainable and must be increasingly being manipulated in the statistics.
And then a huge exchange-rate crisis occurred, and well-known major projects were suddenly exposed to be utterly insolvent and covered-up by corrupt officials, etc. etc. Cue explosively contagious crisis that engulfed even countries with apparently good behaviour, with low debt and non-grossly-manipulated exchange rates.
But within a few years it became clear that growth was resuming, and that the insolvency of projects was linked more to boring old small-scale self-serving corruption rather than a deep fundamental macroeconomic truth uncovered by the Solow model. It had been just a temporary panic, rather than a revelation of doomed stagnation. Oh well.
Btw, @Goumindong it seems like you were looking to fix the variance a bit too much for a game of incomplete information. Trade wars are a bitch, but them and cyber warfare are the highest earning means of dispute in the current "free trade" governance.
You need to figure out where you wanna draw your abstracts. On the value or on the information
for the relevant levels of debt we are concerned about, mainstream macro generally argues that panic or lack thereof tends to be self-reinforcing. Sovereign debt default has a bank-run-esque dynamic here: if a lot of investors pull out, then the state will have trouble rolling over the debt, and so it makes sense for even more investors to pull out. This is argued to be why solvency crises tend to be sudden.
It can be nothing more than a panic: countries can, and have, simply suppressed it through capital controls (to prevent frightened investors from moving their funds out of the country, so that they have nothing else to buy with the currency) and IMF emergency loans (i.e., resorting to the lender of last resort) and then a few years on, bond repayments continue successfully, and it becomes apparent that there was no fundamental insolvency. The controls might then be lifted and the IMF loans repaid.
given such a dynamic, the reason can be pretty much anything.
I guess what I find interesting about all this furor over the validity of the 90% debt limit is, since basically anything can cause investors to panic, the appearance of some kind of debt limit could act as an actual debt limit, if the reporting and dissemination of the limit is strong enough. I think investors are probably more scared of the idea of abandoning american debt than doing anything about exceeding 90% in the case of the states, but if it had been relevant to another country, could that have been enough to cause the kind of panic you were describing? It just seems a little surreal.
I think so.
Back during the 1997 Asian financial crisis there was actually a vaguely similar theory floating about - some econometricians (Alywn Young, etc.) had studied the contributors of economic growth in the Asian tigers and found that much of the growth had appeared to come from adding more and more labour and capital (pulling women from households to factories etc.) rather than increasing the productivity of each added unit of labour and capital. The argument was thus that the growth would have to abruptly slow down once they ran out of women to shift to factories, and so the high predicted growth rates weren't really justified and that a huge banking crisis was going to occur once new investments turn out to be less profitable than expected. See, e.g., The Myth of Asia's Miracle, published in Foreign Affairs in 1994 by one of the world's most influential international trade economists, a certain Paul Krugman (yeah :rotate: . Note that Krugman himself correctly diagnosed the crisis when it then hit, although many of the theory's supporters did not). The intimation was that, like the Soviet Union, the pace of growth was straightforwardly unsustainable and must be increasingly being manipulated in the statistics.
And then a huge exchange-rate crisis occurred, and well-known major projects were suddenly exposed to be utterly insolvent and covered-up by corrupt officials, etc. etc. Cue explosively contagious crisis that engulfed even countries with apparently good behaviour, with low debt and non-grossly-manipulated exchange rates.
But within a few years it became clear that growth was resuming, and that the insolvency of projects was linked more to boring old small-scale self-serving corruption rather than a deep fundamental macroeconomic truth uncovered by the Solow model. It had been just a temporary panic, rather than a revelation of doomed stagnation. Oh well.
the pulling women from households to factories kind of reminds me of the elizabeth warren talk on the shrinking middle class and rise of two-income families that was linked last page - I'd never seen that before and it was pretty interesting. It makes me wonder if changes in who works or what people earn in real terms could be a permanent change in the first world. I remember Krugman had a post on his blog a month or two ago about automation and robotics possibly leading to there being highly capital-intensive (but not labour-intensive) investment opportunities that might crowd out traditional lower paying jobs - the most obvious example I can think of is cashiers in retail, I see those machines all the time and think they're great, but there are probably negative consequences too. There are also quite a few skilled/educated jobs that can be at least partially automated. It seems like this would cause the demand for labour to go down?
for the relevant levels of debt we are concerned about, mainstream macro generally argues that panic or lack thereof tends to be self-reinforcing. Sovereign debt default has a bank-run-esque dynamic here: if a lot of investors pull out, then the state will have trouble rolling over the debt, and so it makes sense for even more investors to pull out. This is argued to be why solvency crises tend to be sudden.
It can be nothing more than a panic: countries can, and have, simply suppressed it through capital controls (to prevent frightened investors from moving their funds out of the country, so that they have nothing else to buy with the currency) and IMF emergency loans (i.e., resorting to the lender of last resort) and then a few years on, bond repayments continue successfully, and it becomes apparent that there was no fundamental insolvency. The controls might then be lifted and the IMF loans repaid.
given such a dynamic, the reason can be pretty much anything.
I guess what I find interesting about all this furor over the validity of the 90% debt limit is, since basically anything can cause investors to panic, the appearance of some kind of debt limit could act as an actual debt limit, if the reporting and dissemination of the limit is strong enough. I think investors are probably more scared of the idea of abandoning american debt than doing anything about exceeding 90% in the case of the states, but if it had been relevant to another country, could that have been enough to cause the kind of panic you were describing? It just seems a little surreal.
I think so.
Back during the 1997 Asian financial crisis there was actually a vaguely similar theory floating about - some econometricians (Alywn Young, etc.) had studied the contributors of economic growth in the Asian tigers and found that much of the growth had appeared to come from adding more and more labour and capital (pulling women from households to factories etc.) rather than increasing the productivity of each added unit of labour and capital. The argument was thus that the growth would have to abruptly slow down once they ran out of women to shift to factories, and so the high predicted growth rates weren't really justified and that a huge banking crisis was going to occur once new investments turn out to be less profitable than expected. See, e.g., The Myth of Asia's Miracle, published in Foreign Affairs in 1994 by one of the world's most influential international trade economists, a certain Paul Krugman (yeah :rotate: . Note that Krugman himself correctly diagnosed the crisis when it then hit, although many of the theory's supporters did not). The intimation was that, like the Soviet Union, the pace of growth was straightforwardly unsustainable and must be increasingly being manipulated in the statistics.
And then a huge exchange-rate crisis occurred, and well-known major projects were suddenly exposed to be utterly insolvent and covered-up by corrupt officials, etc. etc. Cue explosively contagious crisis that engulfed even countries with apparently good behaviour, with low debt and non-grossly-manipulated exchange rates.
But within a few years it became clear that growth was resuming, and that the insolvency of projects was linked more to boring old small-scale self-serving corruption rather than a deep fundamental macroeconomic truth uncovered by the Solow model. It had been just a temporary panic, rather than a revelation of doomed stagnation. Oh well.
the pulling women from households to factories kind of reminds me of the elizabeth warren talk on the shrinking middle class and rise of two-income families that was linked last page - I'd never seen that before and it was pretty interesting. It makes me wonder if changes in who works or what people earn in real terms could be a permanent change in the first world. I remember Krugman had a post on his blog a month or two ago about automation and robotics possibly leading to there being highly capital-intensive (but not labour-intensive) investment opportunities that might crowd out traditional lower paying jobs - the most obvious example I can think of is cashiers in retail, I see those machines all the time and think they're great, but there are probably negative consequences too. There are also quite a few skilled/educated jobs that can be at least partially automated. It seems like this would cause the demand for labour to go down?
quite. It is for this reason that people work shorter hours than they did at the height of the industrial revolution, but what work they can get during the shorter hours is generally much high-paying, thanks to the new machinery.
You're going to have to be a bit more specific when you @ someone in a 2 year old 80 page thread
Sorry, was talking about your last post and I get fucked trying to quote on my phone. I was just reading the most recent page.
My last post was about D/Y tolerances and doesn't (seem) have to do with the stuff you were talking about so I am still confused
Sorry, yea, my wording was obtuse from typing with thumbs, and one of the main reasons I quit bothering with these more heady chats.. I just didn't think I was being confrontational, it's just the way of viewing the Phillips curve. Due to the latency in data for unemployment, it's the infrastructure to know where you're at. With a global marketplace, non-State employment isn't directly controlled, so can only be encouraged to prosper, much like crops. The issue is the zero sum nature of it's current iteration, with no true global infrastructure initiatives. Seeing the politics of the Euro is very relevant.
You're going to have to be a bit more specific when you @ someone in a 2 year old 80 page thread
Sorry, was talking about your last post and I get fucked trying to quote on my phone. I was just reading the most recent page.
My last post was about D/Y tolerances and doesn't (seem) have to do with the stuff you were talking about so I am still confused
Sorry, yea, my wording was obtuse from typing with thumbs, and one of the main reasons I quit bothering with these more heady chats.. I just didn't think I was being confrontational, it's just the way of viewing the Phillips curve. Due to the latency in data for unemployment, it's the infrastructure to know where you're at. With a global marketplace, non-State employment isn't directly controlled, so can only be encouraged to prosper, much like crops. The issue is the zero sum nature of it's current iteration, with no true global infrastructure initiatives. Seeing the politics of the Euro is very relevant.
There's a lot of words here, but I have no idea what you are saying...
e: an attempt at translation: because we have bad lags in gathering employment data (true), states rely on other policy instruments to achieve low unemployment rather than engineering employment directly (also true as an observation, but not quite related to the lag problem). these policy instruments are presently zero-sum-ish (kinda true, as far as monetary policy objectives go).
I have no idea what "looking to fix the variance a bit too much for a game of incomplete information. Trade wars are a bitch, but them and cyber warfare are the highest earning means of dispute in the current 'free trade' governance." means though.
At the most base level they should know that if they're trying to find the debt level cutoff for nations and theor theory thinks they should all be relatively different that averaging cannot give any evidence that a cutoff is universal. They would have to have been looking at the individual data for all countries in order to make that claim. Otherwise they have to say "this is the level for everyone but the UK" or whatever
This was the line I was referring to, I don't wanna quote outta context. I can write about how the means are unique in each case, but I haven't read the whole thread and don't wanna cover previously covered stuff/waste my time. I was just saying with regards to unemployment, the definition of a sovereign actor isn't quite what it used to be.
You're not being confrontational I just don't know what you're talking about. I don't know what variance I am fixing I don't know what game we are talking about I don't know what trade wars or cyberwar has to do with the traditional Phillips curve.
As I understand it the Phillips curve is a short term employment/inflation trade off which more or less falls out of the static IS/LM framework though it does so in an imprecise way. IE take IS/LM and assume there exists constant price level change over time and similar unemployment. Then we see a price level change above that it must be the case that unemployment has fallen.
Basically every sentence of your last post makes sense as a stand alone phrase. But they aren't making sense to me regarding the topic at hand or in conjunction with each other. Maybe I a just not familiar enough with the literature but that is the issue I am seeing.
Are you saying that the Phillips curve is a product of data latency? Are you saying it doesn't exist? Are you saying that static misery coefficients are wrong? Are you saying that because the mechanism for the Phillips curve isn't direct we ignore it?
My point about the Phillips curve was that the identification of the curve was not visual and was both based on theory with statistics confirming the correlation. And that the R/R working paper does not harem to have even this basic framework attached. IE I was making the claim that the working paper as written was not as scientifically rigorous as the Phillips curve work even if at one point we had said "look at this Phillips curve".
At the most base level they should know that if they're trying to find the debt level cutoff for nations and theor theory thinks they should all be relatively different that averaging cannot give any evidence that a cutoff is universal. They would have to have been looking at the individual data for all countries in order to make that claim. Otherwise they have to say "this is the level for everyone but the UK" or whatever
This was the line I was referring to, I don't wanna quote outta context. I can write about how the means are unique in each case, but I haven't read the whole thread and don't wanna cover previously covered stuff/waste my time. I was just saying with regards to unemployment, the definition of a sovereign actor isn't quite what it used to be.
Basically the quoted section isnt talking about unemployment it's talking about GDP growth. Rogoff et all are making the claim that 90% D/Y is the point at which a structural break occurs for roughly all nations. I am saying that the method that they used to identify that break does not work.
This is because their theory suggests that you should calculate the break point for each nation and then look at that distribution but what they did was average the growth of each nation and generate the breakpoint for the average growth. Averaging the growth of each nation cannot identify what they want to identify. It will generate a number but that number won't mean anything
and I am saying, yes, it doesn't work, but it's how policy thinkers thought of an identical D/Y across countries before R & R anyway. In fact, thinking of fixed D/Y at all is a little silly because it ignores the composition of what the D is being raised to spend (think funding social security vs funding long-term infrastructure projects vs military adventures) and how much of it could realistically be subject to emergency levies or capital controls, or even if the government is capable of doing so as a matter of political will and bureaucratic infrastructure. that's the intuition behind "how much can you really tax in a hurry?" in talking about D/Y, after all.
and yet policymakers still wrote down D/Y limits. In between our current state of relative ignorance and the happy, distant future where we understand the drivers of sovereign debt default, we still have to be able to make decisions of how much D/Y is "safe" and that's why the apparent non-rigorous discovery of average growth falling off a cliff at >0.9 D/Y was so convenient
I'm gonna need to do some digging in my backlog of reading to do some citations.. I definitely get where the disconnect is coming from and just gotta explain it a bit better. I went on a bit of a tangent tying it to the war just because of me loving to obfuscate things. I just jump on that, since it's the US reliance on traditional means of warfare that've gotten it into the quagmire, and it's just starting to realize it. Let me find the exact paper I wanted to cite, it was just one I remembered reading a while back when I was reading a lot of economic stuff.
And I agree 100% about the power of 'statistical bias', it's very much in human nature to find non-existent patterns in raw data. It's gonna be a while before I can find the paper, I think it might've been one from Krugman himself in the NYT even. I'll have to do some googling and see if I can find the one, I just can't remember enough particulars/don't have archive access to NYT.
and yet policymakers still wrote down D/Y limits. In between our current state of relative ignorance and the happy, distant future where we understand the drivers of sovereign debt default, we still have to be able to make decisions of how much D/Y is "safe" and that's why the apparent non-rigorous discovery of average growth falling off a cliff at >0.9 D/Y was so convenient
.9 was conveniently close. .9 served to both wave away the problems of past spending while giving a club to go after new spending.
.9 allowed people who wanted to be "serious" to feel like they had some data to back up their fear of big numbers.
it met a number of heuristic and confirmation bias based short cuts for people who wanted to see themselves
as the solvers of the "long terms" problem point to something they could solve.
Seruko on
"How are you going to play Dota if your fingers and bitten off? You can't. That's how" -> Carnarvon
"You can be yodeling bear without spending a dime if you get lucky." -> reVerse
"In the grim darkness of the future, we will all be nurses catering to the whims of terrible old people." -> Hacksaw
"In fact, our whole society will be oriented around caring for one very decrepit, very old man on total life support." -> SKFM
I mean, the first time I met a non-white person was when this Vietnamese kid tried to break my legs but that was entirely fair because he was a centreback, not because he was a subhuman beast in some zoo ->yotes
Posts
Except this doesn't happen in the real world. Consumers don't even take into account taxes they have to pay at the register in 30 seconds, let alone tax adjustments they need to make for effects 10 years hence.
There are no rational actors in the market.
"You can be yodeling bear without spending a dime if you get lucky." -> reVerse
"In the grim darkness of the future, we will all be nurses catering to the whims of terrible old people." -> Hacksaw
"In fact, our whole society will be oriented around caring for one very decrepit, very old man on total life support." -> SKFM
I mean, the first time I met a non-white person was when this Vietnamese kid tried to break my legs but that was entirely fair because he was a centreback, not because he was a subhuman beast in some zoo ->yotes
Ideally yes as far as I can tell. They probably would not need to do a full VAR if they made some assumptions but it probably would not hurt.
If you have panel data there isn't really much reason to not use it as a panel. In the case of this paper it looks as if they didn't do a regression on their data at all. And in addition they did write a conclusion as if their work had value. That is very suspect since you should not have conclusions if you haven't done the work yet. To me it looks like they did a data description and called it good. That being said I only see the working paper at home which may not contain more work which exists in the published version(was r/r published?).
Still looks pretty damning to me. If I were reviewing that piece as I see it I would not accept it for publication in its current state. Like seriously that is low undergraduate quality. They never even state their model.
I cannot believe that this is what the austerity folks got hung up on
Edit Krugman in his next post is discussing regression(i think) results so it may be the case that what we are seeng is missing the crucial price of the pie
in the book they went and did the the event studies and such, it seems. the paper itself was not intended to be rigorous. Apparently it was even less than not-rigorous, though.
e:
that's not really fair, it's basically phillips-curve-type writing. "hey look, what a weird apparently-structural relationship that our book may explain. Now buy/cite our book. thanks."
a little scummy but it's a common tactic and the model is in the book.
None of that should really be kosher in a field that aspires to call itself science.
That's using your conclusion to find evidence, rather than the other way around. Kind of a big deal.
http://s3-ec.buzzfed.com/static/enhanced/webdr06/2013/4/9/12/anigif_enhanced-buzz-17086-1365524052-3.gif
As influential as this book apparently is, how in the world does it take almost 4 years to even give the numbers a quick once over?
I mean, who would think that to do it you need to excise certain cells in a spreadsheet?
Fair point.
I just can't believe how blatantly unethical this has been. I mean, compare this to the "No Global Warming, We Promise" guys that were shouted down by the scientific community repeatedly and continually. It takes 5 minutes for an average person to hear both sides of that argument, do a bit of googling, and find out who's been paid off.
"It is difficult to get a man to understand something, when his salary depends upon his not understanding it!"
"You can be yodeling bear without spending a dime if you get lucky." -> reVerse
"In the grim darkness of the future, we will all be nurses catering to the whims of terrible old people." -> Hacksaw
"In fact, our whole society will be oriented around caring for one very decrepit, very old man on total life support." -> SKFM
I mean, the first time I met a non-white person was when this Vietnamese kid tried to break my legs but that was entirely fair because he was a centreback, not because he was a subhuman beast in some zoo ->yotes
That is fair, but i was reading it outside having read their book. The methodology in the paper doesn't identify the stylized fact it just describes the data. I.E. it has motivation to run the run the data with a particular model but doesn't describe the model they're motivated to use or the results of the test.
I mean i can look at a data set and say "wow it looks like there is a regime change right here" but unless i test that i haven't identified a regime change.
the problem is that R/R don't have a model that gives any reason to believe that there is a regime change that applies for all countries and all years in vastly different circumstances
what they had, and have, is a model that gives regime changes at different cutoffs for some regions and others. their earlier non-bullshit work basically made the well-duh argument that countries with high risk premia due to serial default tended to continue to have high risk premia for loooong times. so you can run regressions to estimate exactly what risk premia level they got stuck at. well, sure, why not, that's hardly surprising. but the model stops there: it just says that countries will have this cutoff that they can't really shift, it doesn't say what the cutoff is
but the later paper is every academic's fantasy: that the approach they discovered is not only generalizable, but generates this astonishing and dramatic result for a whole swathe of other data. it's really super convenient:
the status quo before R/R was that nobody knew what level of debt burden is sustainable or tolerable, clearly the naive OLG optimal-debt-level models aren't plausible given the tendency for stuff to explode for no apparent reason. So it's actually pretty dang amazing! Not so amazing that it is unbelievable - everyone concedes that there is some level of D/Y where it becomes incredible that anyone who buys in will be repaid. maybe advanced industrial economies all do wind up with the same-ish cutoff for structural reasons, who knows? That would certainly be a highly elegant outcome.
it's not: countries will have this cutoff wrt debt load. It's just: countries will have a cutoff, and gosh it does seem to be the same-ish cutoff, amazing.
like I posted earlier, this is "look at the Phillips curve! Look at it!" level of argument. You can't say that five decades of looking very carefully at Phillips curves hasn't been highly impactful; nobody needs to contend that all countries face the same tradeoff, or even that the tradeoff is structural, to claim that it is relevant to policymaking
At the most base level they should know that if they're trying to find the debt level cutoff for nations and theor theory thinks they should all be relatively different that averaging cannot give any evidence that a cutoff is universal. They would have to have been looking at the individual data for all countries in order to make that claim. Otherwise they have to say "this is the level for everyone but the UK" or whatever
also, how do you tell that you have gone over an acceptable amount of debt? can't the things (I guess bond prices would plummet, or something) that you might ascribe to having unacceptable debt levels also be blamed on other things?
when you have any non-trivial amount of debt, you have to have to be able to roll it over: to issue new bonds to fund paying off earlier bonds. this is fine because governments are not households with a finite expected lifespan.
but rolling over the bonds requires investors who believe that they, too, will be repaid in inflation-adjusted terms when their time comes. if you can't convince them to keep buying, then you've run up too much debt, because now you owe far more money than you can plausibly raise in emergency taxes and levies. having your own currency doesn't help either, because whilst you can satisfy the nominal values of your debts, everyone realizes that you failed to repay in real terms, and treats you the same as if you had just defaulted
What I meant by a limit being arbitrary is, as you say, the investors who previously were buying debt for a given country would have to stop buying that debt in order for negative consequences to manifest. They have to have a reason for deciding to stop buying debt. What constitutes a reason, for a western economy that is basically stable/isn't facing any special circumstances that might add risk?
for the relevant levels of debt we are concerned about, mainstream macro generally argues that panic or lack thereof tends to be self-reinforcing. Sovereign debt default has a bank-run-esque dynamic here: if a lot of investors pull out, then the state will have trouble rolling over the debt, and so it makes sense for even more investors to pull out. This is argued to be why solvency crises tend to be sudden.
It can be nothing more than a panic: countries can, and have, simply suppressed it through capital controls (to prevent frightened investors from moving their funds out of the country, so that they have nothing else to buy with the currency) and IMF emergency loans (i.e., resorting to the lender of last resort) and then a few years on, bond repayments continue successfully, and it becomes apparent that there was no fundamental insolvency. The controls might then be lifted and the IMF loans repaid.
given such a dynamic, the reason can be pretty much anything.
Where does one start to learn some of this stuff?
I teach college economics for a living. To the interested individual, I always recommend the following book. Macroeconomic Essentials. It is clear, short, and doesn't require a big math background. It also helps that it isn't very expensive. At the same time it manages to be rigorous. It's one of those books that you can understand the first time through, yet every time you read it you discover another insight that you didn't appreciate the first time.
Now the specific thing we were just discussing is closer to a topic called 'Econometrics.' This is the intersection of economic models and how you implement them on statistical data. There is an excellent book I can recommend on this subject that has gotten many an economist through their Ph.D. program, yet can be read by non-economists. A Guide to Econometrics.
You will notice that it is written by the same author. That's no accident. Peter Kennedy is the kind of economist who understands his field so well, he can explain even difficult concepts with crystal clarity. I read this Econometrics text in grad school. Because of how good it was I looked around what other textbooks he had written and found his macro book that way. I even taught out of it for a while. But I have found that it's better for studying on your own. For a course, students do better with a more comprehensive book that has problem sets, supporting websites, online homework problems, and all that jazz. I use N. Gregory Mankiw these days, same as just about everyone else. You know, this guy:
http://media.mtvnservices.com/embed/mgid:cms:item:comedycentral.com:362047
I guess what I find interesting about all this furor over the validity of the 90% debt limit is, since basically anything can cause investors to panic, the appearance of some kind of debt limit could act as an actual debt limit, if the reporting and dissemination of the limit is strong enough. I think investors are probably more scared of the idea of abandoning american debt than doing anything about exceeding 90% in the case of the states, but if it had been relevant to another country, could that have been enough to cause the kind of panic you were describing? It just seems a little surreal.
I think so.
Back during the 1997 Asian financial crisis there was actually a vaguely similar theory floating about - some econometricians (Alywn Young, etc.) had studied the contributors of economic growth in the Asian tigers and found that much of the growth had appeared to come from adding more and more labour and capital (pulling women from households to factories etc.) rather than increasing the productivity of each added unit of labour and capital. The argument was thus that the growth would have to abruptly slow down once they ran out of women to shift to factories, and so the high predicted growth rates weren't really justified and that a huge banking crisis was going to occur once new investments turn out to be less profitable than expected. See, e.g., The Myth of Asia's Miracle, published in Foreign Affairs in 1994 by one of the world's most influential international trade economists, a certain Paul Krugman (yeah :rotate: . Note that Krugman himself correctly diagnosed the crisis when it then hit, although many of the theory's supporters did not). The intimation was that, like the Soviet Union, the pace of growth was straightforwardly unsustainable and must be increasingly being manipulated in the statistics.
And then a huge exchange-rate crisis occurred, and well-known major projects were suddenly exposed to be utterly insolvent and covered-up by corrupt officials, etc. etc. Cue explosively contagious crisis that engulfed even countries with apparently good behaviour, with low debt and non-grossly-manipulated exchange rates.
But within a few years it became clear that growth was resuming, and that the insolvency of projects was linked more to boring old small-scale self-serving corruption rather than a deep fundamental macroeconomic truth uncovered by the Solow model. It had been just a temporary panic, rather than a revelation of doomed stagnation. Oh well.
Btw, @Goumindong it seems like you were looking to fix the variance a bit too much for a game of incomplete information. Trade wars are a bitch, but them and cyber warfare are the highest earning means of dispute in the current "free trade" governance.
You need to figure out where you wanna draw your abstracts. On the value or on the information
You're going to have to be a bit more specific when you @ someone in a 2 year old 80 page thread
the pulling women from households to factories kind of reminds me of the elizabeth warren talk on the shrinking middle class and rise of two-income families that was linked last page - I'd never seen that before and it was pretty interesting. It makes me wonder if changes in who works or what people earn in real terms could be a permanent change in the first world. I remember Krugman had a post on his blog a month or two ago about automation and robotics possibly leading to there being highly capital-intensive (but not labour-intensive) investment opportunities that might crowd out traditional lower paying jobs - the most obvious example I can think of is cashiers in retail, I see those machines all the time and think they're great, but there are probably negative consequences too. There are also quite a few skilled/educated jobs that can be at least partially automated. It seems like this would cause the demand for labour to go down?
quite. It is for this reason that people work shorter hours than they did at the height of the industrial revolution, but what work they can get during the shorter hours is generally much high-paying, thanks to the new machinery.
Sorry, was talking about your last post and I get fucked trying to quote on my phone. I was just reading the most recent page.
My last post was about D/Y tolerances and doesn't (seem) have to do with the stuff you were talking about so I am still confused
Sorry, yea, my wording was obtuse from typing with thumbs, and one of the main reasons I quit bothering with these more heady chats.. I just didn't think I was being confrontational, it's just the way of viewing the Phillips curve. Due to the latency in data for unemployment, it's the infrastructure to know where you're at. With a global marketplace, non-State employment isn't directly controlled, so can only be encouraged to prosper, much like crops. The issue is the zero sum nature of it's current iteration, with no true global infrastructure initiatives. Seeing the politics of the Euro is very relevant.
There's a lot of words here, but I have no idea what you are saying...
e: an attempt at translation: because we have bad lags in gathering employment data (true), states rely on other policy instruments to achieve low unemployment rather than engineering employment directly (also true as an observation, but not quite related to the lag problem). these policy instruments are presently zero-sum-ish (kinda true, as far as monetary policy objectives go).
I have no idea what "looking to fix the variance a bit too much for a game of incomplete information. Trade wars are a bitch, but them and cyber warfare are the highest earning means of dispute in the current 'free trade' governance." means though.
This was the line I was referring to, I don't wanna quote outta context. I can write about how the means are unique in each case, but I haven't read the whole thread and don't wanna cover previously covered stuff/waste my time. I was just saying with regards to unemployment, the definition of a sovereign actor isn't quite what it used to be.
As I understand it the Phillips curve is a short term employment/inflation trade off which more or less falls out of the static IS/LM framework though it does so in an imprecise way. IE take IS/LM and assume there exists constant price level change over time and similar unemployment. Then we see a price level change above that it must be the case that unemployment has fallen.
Basically every sentence of your last post makes sense as a stand alone phrase. But they aren't making sense to me regarding the topic at hand or in conjunction with each other. Maybe I a just not familiar enough with the literature but that is the issue I am seeing.
Are you saying that the Phillips curve is a product of data latency? Are you saying it doesn't exist? Are you saying that static misery coefficients are wrong? Are you saying that because the mechanism for the Phillips curve isn't direct we ignore it?
My point about the Phillips curve was that the identification of the curve was not visual and was both based on theory with statistics confirming the correlation. And that the R/R working paper does not harem to have even this basic framework attached. IE I was making the claim that the working paper as written was not as scientifically rigorous as the Phillips curve work even if at one point we had said "look at this Phillips curve".
Basically the quoted section isnt talking about unemployment it's talking about GDP growth. Rogoff et all are making the claim that 90% D/Y is the point at which a structural break occurs for roughly all nations. I am saying that the method that they used to identify that break does not work.
This is because their theory suggests that you should calculate the break point for each nation and then look at that distribution but what they did was average the growth of each nation and generate the breakpoint for the average growth. Averaging the growth of each nation cannot identify what they want to identify. It will generate a number but that number won't mean anything
and yet policymakers still wrote down D/Y limits. In between our current state of relative ignorance and the happy, distant future where we understand the drivers of sovereign debt default, we still have to be able to make decisions of how much D/Y is "safe" and that's why the apparent non-rigorous discovery of average growth falling off a cliff at >0.9 D/Y was so convenient
http://www.bkconnection.com/static/Building_a_Win-Win_World_EXCERPT.pdf
.9 was conveniently close. .9 served to both wave away the problems of past spending while giving a club to go after new spending.
.9 allowed people who wanted to be "serious" to feel like they had some data to back up their fear of big numbers.
it met a number of heuristic and confirmation bias based short cuts for people who wanted to see themselves
as the solvers of the "long terms" problem point to something they could solve.
"You can be yodeling bear without spending a dime if you get lucky." -> reVerse
"In the grim darkness of the future, we will all be nurses catering to the whims of terrible old people." -> Hacksaw
"In fact, our whole society will be oriented around caring for one very decrepit, very old man on total life support." -> SKFM
I mean, the first time I met a non-white person was when this Vietnamese kid tried to break my legs but that was entirely fair because he was a centreback, not because he was a subhuman beast in some zoo ->yotes
my conclusion after one minute of glancing: this is pseudoscientific babble
and I say that given the low, low bar I have granted for mainstream macroeconomics as a science above, too