EC101: The fallacy of composition
I get tired of correcting Keynesians who don’t understand how to estimate the multiplier. But as long as they keep saying things like this, I’ll have to keep doing so:
In general, cross-sectional comparisons are proving to be a very good way to test some propositions in macroeconomics. I’d cite, for example, the Nakamura-Steinsson paper (pdf) that uses fluctuation defense spending “” which has very unequal impacts across states “” to estimate the multiplier on fiscal policy (it’s about 1.5).
Actually cross-section studies are a lousy way to test for the multiplier, as they suffer from the fallacy of composition. I’ve pointed this out many times, but Paul Krugman obviously doesn’t read my blog, otherwise he wouldn’t keep making these elementary errors. Of course he’s told us that he doesn’t like to read conservative blogs because they contain nothing of value.
Then there is this, from the very same post:
Calculated Risk sends us to two papers by Amir Sufi and Atif Mian using county-level data to investigate the causes of the recession. Their work strongly supports the balance-sheet view: a fall in demand from highly indebted households is the big story, and you can confirm that by showing that the big declines in nontradable employment “” that is, employment in industries that sell locally “” is in those countries where debt levels were high.
And let me guess, counties with economies dominated by autos and steel usually suffer bigger job losses in recessions than counties dominated by hospitals and colleges. Does that tell us anything about what causes recessions? Doesn’t EC101 also teach that correlation doesn’t prove causation?
Let’s suppose recessions were caused by tight money, not balance sheet problems. And let’s suppose that tight money reduces nominal income. And let’s suppose that most debts are nominal, not indexed to inflation. In that case wouldn’t you expect tight money to lead to bigger spending declines in highly indebted areas, even if debt played no role in causing the recession?
Fallacy of composition. . .
Correlation doesn’t prove causation. . .
Krugman’s textbook must have something to say on those topics.
Tags:
17. November 2011 at 17:23
Excellent post. You’ve said it before but it needed to be said again.
Partial equilibrium is not general equilibrium and cross-sectional variation is a really lousy way of trying to infer general equilibrium effects.
I still find these “the multiplier is 1.5” stories mysterious. Is “the multiplier” some kind of fundamental constant of the universe? Even if one were able to tease out some sort of multiplier at the margin, how useful would that number be for evaluating policies that are certainly not marginal — like an expansion of fiscal purchases by $800bn or $1 trillion?
17. November 2011 at 18:18
Once computing became easy and inexpensive and data easily available, researches can “validate” any “prejudice” they embrace!
17. November 2011 at 18:44
Off-subject, but NY Fed President Bill Dudley did a nice interview with PBS Nightly Business Report today, calling for more asset purchases and a more explicit forward-looking target to get asset prices to rise via the expectations channel, etc. You have to overlook the comments calling for more expansive fiscal policy, but it’s worth a watch/read.
17. November 2011 at 18:51
@JTapp
How many times, in the last several months we´ve heard that the Fed “is not out of ammo”? It´s becoming a big and dangerous joke!
http://thefaintofheart.wordpress.com/2011/11/17/we-keep-hearing/
17. November 2011 at 19:46
I don’t know Scott. You grouse that Krugman doesn’t read you. But do you read this?
http://diaryofarepublicanhater.blogspot.com/2011/11/i-try-to-get-to-bottom-of-sumner-kimel.html
For my part I would agree witb Krugman that watching Beck or listening to Limbaugh is a waste of time. I don’t feel that way about you if you’re a conservative blog then I read a conservative blog. To an extent economics are different than politics. Not much gets said in politics by self identified conservatives that I have much respect for. In economics there are some very good ones that are conservatives.
17. November 2011 at 21:14
Negative interest rates?
http://economistsview.typepad.com/economistsview/2011/11/why-hasnt-the-fed-lowered-the-rate-it-pays-on-reserves.html
17. November 2011 at 21:21
While I never find much worthy of commentary in any of Krugman’s posts the one thing I do find intriguing is that he notes is the “balance sheet recession.”
To direct proper policy measures it is painfully apparent that we need to first understand the problem. Of course Keynesian’s have been far more interested in trying to prove the existence of the mythical multiplier vice grasping our current dilemma.
17. November 2011 at 21:30
Krugman seems to have two claims in that post. Given that the fall in spending is disproportionately concentrated among those affected by the collapse of housing values, he claims:
1. Blaming the recession and continuing unemployment on “structural” economic issues or business uncertainty is idiotic.
2. The key cause of the recession and continuing unemployment is the worsening of household balance sheets.
Do you disagree with him on both claims, or only on claim #2?
17. November 2011 at 21:49
I took a class at UCLA where we used Krugman’s introductory micro textbook. I thought it was funny how he was obligated to put a graph in explaining taxation’s negative effect on production. That graph completely obliterated everything he stood for and you could tell he was extremely uncomfortable writing about it.
18. November 2011 at 03:36
“Actually cross-section studies are a lousy way to test for the multiplier, as they suffer from the fallacy of composition. I’ve pointed this out many times….”
You have asserted it repeatedly, yes. However, argument by repeated assertion is unconvincing. Until you actually show your work, why should I believe you?
18. November 2011 at 06:27
@Kevin Donoghue:
He has shown his work. Your argument that he hasn’t is mere assertion. It’s simple logic, and the same logic that applies to the Keynesian “paradox of thrift.”
If you tax A, B, and C to give to A, determining that A is relatively better off than B and C afterwards doesn’t prove that A, B, and C as a whole are better off.
The cross-section technique described is silly, and if used overestimate *any* sort of “multiplier.” You can demonstrate that tax cuts have an enormous multiplier by cutting the taxes of only some people, and then observing that places which got more taxes cut are doing better than those that didn’t.
Would Krugman believe a study that “proves” that the paradox of thrift doesn’t exist by noting that people that save are better off than those who don’t, therefore the macroeconomic effect of savings *must* be positive?
Krugman does believe in the paradox of thrift– see search results. So he’s simply being illogical and ideological, as he does too often.
I don’t enjoy listening to Beck or Limbaugh for the same reasons that Krugman tires me.
18. November 2011 at 06:29
CBO says Stimulus was bad, bad, bad…
http://reason.com/blog/2011/11/17/cbo-on-the-stimulus
18. November 2011 at 06:31
Larry:
If the argument is that the market interest rate would be negative, then that’s definitely a sign of tight money. I don’t see how people can think that the idea that the market would rather have a smaller nominal amount of money later than a given amount now as anything other than deflation and tight money.
Yet people persist in thinking that low interest rates means easy money. Not necessarily.
18. November 2011 at 06:32
So Larry’s linked article, far from a defense of the Fed, is an indictment.
Morgan:
The CBO said that the stimulus was short-term positive but long-term slightly negative. Not the sort of report that’s going to change any minds, and again the sort of thing that’s driven by the CBO’s model, not by much else.
Not that I favored fiscal stimulus, and especially not that terribly designed fiscal stimulus.
18. November 2011 at 06:35
Kevin,
Let’s focus just on the “multiplier” paper that Krugman cites, because it’s particularly suspect.
Mantra: partial equilibrium is not general equilibrium. Suppose the Federal government spends $50bn on a new defense project in California. That certainly raises employment in California and indeed might have some sort of “spillover” or “multiplier” effect, raising economic activity by more than $50 million. However, there are at least four things wrong with trying to tease out “the multiplier” from that data:
1) California gains jobs but other states lose them. So while dumping a lot of Federal money into a single state might increase employment in that state and generate a multiplier above unity, there is no reason to infer that if you dump Federal money into every state, then every state will see a multiplier above unity. PE is not GE.
2) The regression techniques used in these papers are unimpressive. Regressing (change in output = a + b*change in government spending + error) doesn’t even begin to address the identification problem. The estimate of “b” is confounded by omitted variable bias at the very least. Ambrosini has more on this in a useful October 6 post.
3) The central bank moves last. The Fed has a path for nominal spending that it was to obtain and if the fiscal authorities try to stimulate beyond that path, the Fed will contract. The multiplier is identically 0 after the central bank moves. This is Scott’s main argument.
4) “The multiplier” isn’t some sort of constant of the universe. At best the multiplier is a declining function of the size of stimulus. Even if the marginal multiplier were 1.5, that doesn’t mean the average multiplier over a $800bn or $1 trillion or $2 trillion stimulus will exceed unity or even zero.
18. November 2011 at 06:41
John Thacker,
You are confusing the issue by bringing in the question of who (if anyone) is made better off by expansionary fiscal policy. It’s not at all illogical to claim, for example, that the multiplier is 1.5 but fiscal expansion makes everyone worse off; a model with those properties need not involve any contradiction.
As to whether Scott has shown his work, I’ll do a thorough search of his posts on this subject just to make sure, but I’m a regular reader and I’ll be quite surprised if it turns out that he has.
18. November 2011 at 06:45
Kevin,
I may be wrong but I was pretty sure he wrote a post laying out his criticisms of the Feyrer/Sacerdote NBER working paper that calculated a multiplier based on state-level data.
18. November 2011 at 06:48
Ah, he wrote two brief (well, brief for Scott) posts on that paper. I’ll only link one since the spam filter blocks posts with more than one link: http://www.themoneyillusion.com/?p=9056
18. November 2011 at 06:52
John Thacker To your comment
“If you tax A, B, and C to give to A, determining that A is relatively better off than B and C afterwards doesn’t prove that A, B, and C as a whole are better off.”
I’d say the way you frame it we don’t know enough to say whether A,B, and C as a whole are or are not better off. I live in NY and there were two women at the train station today discussing how it’s really tough moving around the city by Wall St right now.
A friend of there’s said her boss told her not to even bother coming in today with all the disorder.
Even if B and C see their taxes raised to give to A there are many scenarios where it could be in their best interest and the interest of the whole. Like if B and C are both part of the top 1 percent of earners and A is part of OWS.
A society with less disorder may well be worth a small tax increase.
By the way this is just one scenario in which it would benefit the whole. It is certainly not the only one.
18. November 2011 at 07:16
“Tight money hurts indebted households,” is one of those logical, indisputable axiom, kind of like, “loose money makes commodity prices spike.”
The implications are interesting: tight money will hurt nominal incomes more in the presence of higher leverage. Which means that the sensitivity of the economy to the path of monetary policy depends on leverage (stay with me). Since the level of leverage depends on the success of the Fed in manufacturing stable NGDP, then:
The more stable NGDP growth, the higher the leverage, the harder it is for the Fed to hit an NGDP target without significant overshoots on both sides.
Ergo: the problem is debt.
18. November 2011 at 07:19
BTW, your argument that
“monetary policy did not raise commodity prices because, look at that chart of China IP and a commodity index,”
is virtually the same as Krugman saying,
“monetary policy did not depress the spending of indebted households because, look at that chart of household leverage vs. spending.”
18. November 2011 at 07:20
Should have read “monetary policy did not..the spending of households,” and not, “indebted households.”
18. November 2011 at 07:25
I’m a new reader of this blog. I have two questions. The first one is related to this post. I would like to know what empirical technology exists that can be used to prove causation. I would vote for event studies that are so common in finance as coming as close as possible to this goal. Unfortunately, the data requirements seem a bit overwhelming if you want to investigate, say, the impact of monetary events from late 2008 to the present. Where would you find oodles of similar events to align in event time?
This brings me to my second question. Can you, Scott, or some reader of this blog point me to a post with its comments on the blog that characterize the posture of monetary policy during the last several years. For example, is the huge expansion of base and monetary aggregates just viewed as temporary?
18. November 2011 at 07:43
See the latest on Sumner vs Krugman and the fallacy of composition http://diaryofarepublicanhater.blogspot.com/2011/11/krugman-vs-sumner-and-fallacy-of.html
18. November 2011 at 08:58
@johnleemk,
Yes, I’ve read both of those posts. The one you linked seems to be the first where Scott claims that there’s a “basic logical error” involved:
“…this study tells us nothing about the macro effects of fiscal stimulus. It’s a near perfect example of the fallacy of composition. Every single anti-stimulus model would predict exactly the same finding at the micro level.”
Now that’s a sweeping statement! But even if it’s true, so what? It doesn’t support the claim made in the previous sentence. A fallacy of composition involves a statement of the form: individual entities can do X, therefore a collection of entities can do X (where in fact adding-up constraints rule that out). Feyrer and Sacerdote make no such error that I can see.
18. November 2011 at 08:59
Thanks Integral.
JTapp, That’s good to hear.
Marcus, That’s right, I think they keep saying that partly because it’s true, and partly because they want to dispel expectations of deflation.
Larry, Thanks for the link.
Mike Sax. I’ve sent Mike Kimel some of my papers, but he really doesn’t understand what happened during the 1930s. Therefore it’s hard for him to analyze the decade, without knowing the stylized facts. By analogy, suppose you were debating the 1990s, and someone kept insisting that there was no high tech stock boom. That the idea that stocks rose sharply was just a myth. It would make it pretty hard to discuss issues with that person, wouldn’t it? I mean the NASDAQ actually did soar to 5000, didn’t it? And the dollar really was devalued sharply between April 1933, to February 1934, wasn’t it? Or do you agree with Kimel, and disgree with every single competent economic historian in the world?
Mike Elliot, The recession wasn’t caused by balance sheet problems, but those problem affected how the recession played out.
I agree 80% with number one, and disagree that balance sheets were a cause of the recession.
John, That’s interesting.
Kevin. In my first post he had linked to a different paper making the same point. I later met one of the the authors of that paper and he admitted that my criticism was valid.
Second, It’s kind of obvious when you think about it. Do you know of a single non-Keynesian model that would deny multplier effects at the local level? I don’t, so what’s this showing? It tells us that if taxpayers from around the country spend $5 billion on a military base in Poduck North Dakota, then the GDP of Podunk North Dakota goes up. Are we supposed to be surprised?
Morgan, That’s good to hear.
John Thacker, Good point about negative equilibrium rates and tight money.
Integral, Thanks for mentioning the Ambrosini post, I forget to mention that. I guess Krugman also doesn’t read Ambrosini.
Kevin, This is my 4th post on the subject, thus I don’t lay out all my objections each time, I’m getting tired of the subject. Keynesians need to stop making these silly claims. Or else stop making fun of silly claims made by Fama, et al.
David, I don’t see any evidence that more leverage makes it harder to hit a NGDP target. But if so, it would also be harder to hit an inflation target or Taylor Rule. I prefer to avoid having the Fed create massive recessions so that Americans are more cautious, and would instead prefer regulation of leverage, if FDIC and TBTF creates a moral hazard market failure.
I don’t see the China analogy. I define monetary policy in terms of US NGDP growth. You might not like that definition, but that’s my definition. So I’m saying the commodity boost wasn’t due to easy money, becuase there is no correlation. So it’s not at all a “correlation doesn’t imply causation” issue. There is no correlation.
I mentioned China (actually all developing countries) to provide an alternative explanation of the movement in commodity prices, which makes more sense (as the developing country growth is commodity intensive, and also fits the facts much better (better correlation.)
18. November 2011 at 09:02
Kevin, By ‘”fallacy of composition” I meant the fallacy of assuming that what’s true for the indivudal case is also true in the aggregate. Good weather for one farm raises that farm’s income. Good weather for all farms often lowers farm incomes. Those sorts of examples are called “the fallacy of composition” in the textbooks.
18. November 2011 at 09:10
Peter, Event stuies are the best (in my view.) That’s where I differ from my fellow macroeconomists, who seem to prefer things like VARs.
In the right margin you’ll see a “link to key blog posts and papers.” That may have what you are looking for. The Fed insists that the big base injection is either temporary, or if permanent will lead to higher future IOR to prevent a breakout of inflation. That’s why the quantity theory predictions haven’t held true.
18. November 2011 at 09:15
Scott,
“I don’t see any evidence that more leverage makes it harder to hit a NGDP target.”
I was using your logic. You said leverage amplifies the effect of tight money; ergo, with leverage, any degree of (passive) tightening manifests itself as a larger decline in nominal spending. This higher sensitivity of spending to policy would make it harder to “fine-tune”, just as if a car steering wheel caused you to veer off the road with just a slight movement.
18. November 2011 at 10:23
Mike Sax,
I think you maybe be retarded.
If you think OWS is much more than an laughable annoyance, you are very likely indeed a tard.
A pissed off Tea Party is HUGE FREAKING DEAL, OWS has no $$, and no votes.
They are entertainment. I’m sad to see them go.
18. November 2011 at 10:26
“Do you know of a single non-Keynesian model that would deny multplier effects at the local level?”
Yes. Consider a simple two-good model: guns and butter. The economy is always at full employment. So multiplier effects are ruled out. The only thing the government does is start a war now and then. Result: more guns, less butter. In a study of the Feyrer-Sacerdote sort, we’d expect to see job gains in gun factories and job losses in creameries.
AFAICT they don’t fall into “the fallacy of assuming that what’s true for the individual case is also true in the aggregate” anywhere in that paper. Yet you keep claiming that they do.
18. November 2011 at 10:32
Could you define “tight money”?
18. November 2011 at 10:52
Kevin,
Correct me if I’m wrong, but their methodology is to calculate how many jobs were added in each state and how much GDP growth was observed in each state, correlate this to federal stimulus spending in each state, and sum up the effects to get a nationwide multiplier — isn’t it? I’m pretty sure this is what Scott is talking about.
18. November 2011 at 10:52
Kevin: in your model, if the government taxed California to buy guns from Wyoming, don’t you think that the GDP of Wyoming would increase? That’s what Scott means by “at the local level”.
18. November 2011 at 10:52
Morgan you really add a lot to an an informed discussion. Calling people retards, what next, you stick out your tongue?
Scott I agree there was a devaluation in 1933-34. As I made clear in my post I’m not entirely clear about Kimel’s argument about this though he says he will update it later.
18. November 2011 at 10:55
I should also add Morgan, that if you read my post, assuming you read at all, I wasn’t necessarily lauding OWS though not simply calling them retards like you.
My point was simply that the social upheaval caused by certain policies may already make them not worth it. To me if I’m a rich guy who works on Wall St. I’d rather pay a 2 percent more in taxes and avoid all this. That was my point. Obviously you are in the wrong class.
18. November 2011 at 11:16
you don’t even understand what “tight money” is. Money is endogenous, it is well known (not to monetarists). People borrow and the Fed follows with reserves. The money is tight because people don’t borrow and banks don’t lend. It is not a cause of the recession, but an effect. I guess Krugman doesn’t read your blog because it is a waste of time, you assert stuff contrary to facts, day in day out.
18. November 2011 at 11:47
good grief, Morgan
18. November 2011 at 12:18
@johnleemk
There is lots of scope for confusion here. Scott’s first attack that I know of along these lines was directed at Feyrer and Sacerdote, whereas he’s now making the same claim about Nakamura and Steinsson. I think he’s wrong in both cases.
@Alex Godofsky
Wyoming’s gain is California’s loss. If there’s a gain without a corresponding loss then initially we were not on the production-possibility frontier, i.e. we could have more guns without giving up any butter.
18. November 2011 at 12:31
@johnleemk
To clarify — I think you’re basically correct as to what Scott is talking about, but it’s up to him to show that the authors have fallen into “the fallacy of assuming that what’s true for the individual case is also true in the aggregate.” It’s no sin to derive nationwide estimates from regional data.
18. November 2011 at 12:35
Kevin: and the entire point is that this is precisely what these studies are doing. They say “Look, Wisconsin got more federal money than California, and it also gained more jobs. Therefore if we give more federal money to every state they will all gain jobs!” But the outcome they observed is completely consistent with theories that predict giving more money to every state will not result in every state gaining jobs.
18. November 2011 at 13:09
Mike Sax,
Yeah, and I think you have very little grip on facts as they exist.
Here’s a fact: http://www.cbo.gov/doc.cfm?index=12485
—-
Please cast your eye on tan line under 1979-2007 Highest Quintile.
Those really are the people that matter, not your annoyed Wall Street Guy.
Those are the 81-99%.
They are the bulk of Tea Party. They own stuff. In every small town across America they own all the small businesses.
Look at their muscle.
In a game of WHO GETS TO KEEP what we shake out of the top 1%, what makes you think it is going to roll down past the 80-99%??
They all vote. They donate. They own stuff. They are pillars of their community.
And THEY WANT to tear down Wall Street more than any idiots hippie child camping out.
Ok, so that’s a far more complete reflection of the state of things.
Let me add a bit more…. the sheer % of Americans who identify with the 81-99% is actually much higher than 1/5, because over their lifetime many Americans touch that space, and they might be in the 61-80%, but they were in the top quintile and they intend to be there again.
Does this help you with my exasperation towards your analysis that giving up 2% more in taxes by the Wall Street crowd isn’t going to get anything for OWS?
I keep saying this: if OWS was a real thing, if they were serious, they would at that chart and say, “the guys we gotta get behind are the Tea Party”
It is the only rational game play someone has against the top 1% – you gotta ADMIT the top half of the 99% ARE IN CHARGE.
18. November 2011 at 13:11
Professor Sumner,
Great work on the blog. I’ve spent a little over an hour reading through some of your posts. I’m very like minded with respect to some of your larger core ideas. I’ve enjoyed reading some of the facts and analysis that you provided. These facts and your analysis have shed some light onto why we are in this position today.
I’m a Bentley grad from the class of ’05. I majored in Economics, but never had the pleasure of having you as a professor. If you run into M. Quinn, tell him I say hello.
18. November 2011 at 13:11
Kevin:
“It’s no sin to derive nationwide estimates from regional data.”
I think you are the one making assertions here without backing them up. Scott’s point is that this is entirely the case — they are deriving nationwide estimates from regional data. It is impossible to argue that because X amount of federal stimulus correlated with 1,000 new jobs in region A and Y amount of federal stimulus correlated with 10,000 new jobs in region B that by necessity, Z amount of federal stimulus will necessarily correlate with some number of new jobs nationwide.
18. November 2011 at 13:41
Again Morgan, I think you personalize too much. I never claimed to be a booster for OWS only pointing out that they’re a symptom of the frustation of most Americans.
You seem to be arguing that they don’t matter but the tea party does. Don’t knoow what’s your basis for that. Is it because they have money? That’s the knock on the tea party that they are financied by the Koches, et al. So your boast is the whole problem thinking those with money count for more.
Or is it because only the people who are in your 81-95 percent matter?
Actually you probably assume too much about the makeup of OWS many of them are in this very same 81-95 percent group. You claim that the tea party better represents the will of the American people but I don’t know what you base it on. To say they have more campaign donations hardly proves they are legitimate.
I’m not as much a cheerleader for OWS as you are for the tea party but I’ll tell you what. If I’m going to go by the cohernence of the demands the OWS protests are more in line with most Americans views. The tea party doesn’t know what it’s for other than seeing Obama’s birth certificate. What now you want his school grades?
We have tea party supporters saying “keep your government hands of my Medicare.” That’s what I mean by incoherence.
18. November 2011 at 13:47
One more point, to your questioin “Does this help you with my exasperation towards your analysis that giving up 2% more in taxes by the Wall Street crowd isn’t going to get anything for OWS?”
I have to say no. You miss what I was saying. If the rich would be willing to make pretty small sacrifices that would be enough to forestall such social disruptions.
It actually would get OWS something significant-it actually represents many billions of dollars of revenue. That’s the whole point. A guy like you fights the idea that the top 1 percent should have to pay anything more.
Meanwhile people like Buffett and Gates are like “we’ll pay it” begging to be taxes a little more.
18. November 2011 at 14:37
Not that it excuses anyone acting like a jerk, but the suggestion that if taxes on the rich were raised 2% the OWS kids would pronounce themselves satisfied and go home is one of the most improbable things I have ever read.
18. November 2011 at 15:03
Well you confuse cause and effect. It was the refusal to do such simple policies orignally that led to OWS. Actually this is what they are effectively demanding right now. What is it that you think they are demanding exactly a holocaust?
I should point out that the the pejorative reference to them as kids can go both ways. One could speak of the tea partiers as old timers.
In a way you make my point in that you find it hard to believe that woudld be enough to satisfy them. But what have they demanded: either allow the Bush tax cuts for the rich to expire or add a new millionaries tax of no more than 5 percent. If you think it is so small a concession why do so many on the Right seem willing to oppose it with as Dick Armey would put it “every fiber of my being.”
Neither of these policies would raise tax rates by that much. Yet it would do a raise revenue.
Meanwhile 68 percent of millionaries themselves are willing to pay more taxes.
“Poll after poll has shown that the American people favor higher taxes on millionaires, while several ultra-wealthy citizens like Liveris, Edwards, and investor Warren Buffett have all said that they should be paying more to support the country. A survey by the Spectrem Group found that “68% of millionaires (those with investments of $1 million or more) support raising taxes on those with $1 million or more in income.”
In fact even ” Dow Chemical CEO Andrew Liveris joined the tax-me-more crowd, saying, “I can tell you the highest taxpayers, including me, should be paying more”
http://thinkprogress.org/
18. November 2011 at 15:18
Jeebus, there’s still someone out there who believes OWS is demanding anything exactly?
18. November 2011 at 15:26
Mike Sax,
AGAIN, the Tea Party should and will be paying less.
1. Any efforts to weaken the grip of the 1% will require the acquiesce of the top 1/3 of America…. the folks who in their lifetime will earn in the 81-99%.
2. They will only agree if they get to keep the lion share of the booty.
Look dude, I know you think the problems facing the OWS crowd are all OUR problems, but they are not.
And if the OWS crowd wants to actually count coup against the top 1%, then you should focus on passing them the knowledge I’m passing on.
Finally, the Tea Party is very simple, they just want government to be smaller. And they are going to win that fight. And the Fed is terrified of them, and therefore will HELP them get what they want – Obama out of office.
Dude, having money is NECESSARY to count. Having votes is also NECESSARY to count.
That’s WHY the Oligarchs (no votes) AND OWS (no money) are doomed to fall to the will of the “silent majority.”
18. November 2011 at 15:32
Here ya go Mike…
http://news.yahoo.com/occupy-memphis-tea-party-members-meet-084306417.html
Its what ALL the OWS guys should be doing… making nice with the guys who matter.
18. November 2011 at 15:52
http://www.frbsf.org/education/activities/chairman/index.html
Scott, Ezra Klein posted on twitter the above link. It is a video game from the StL fed in which you manage policy. I have messed around with it a bit and if you ZLB during a recession, it says that the govt has to step in to encourage expansion. Perhaps letting us into minds of Fed people?
Check it out…
18. November 2011 at 15:57
@Alex Godofsky: “the outcome they observed is completely consistent with theories that predict giving more money to every state will not result in every state gaining jobs.”
The two-good model I mentioned (in response to a question from Scott) makes a stronger prediction than that. It predicts that government spending which boosts employment in one sector will force a contraction in the other; more guns, less butter.
@johnleemk: “I think you are the one making assertions here without backing them up.”
The only assertion I’m making is that Scott hasn’t made a case. Surely the onus isn’t on me? He’s the one accusing various economists of making an ‘EC101’ error.
“Scott’s point is that this is entirely the case “” they are deriving nationwide estimates from regional data.”
And why not? The nation is made up of regions. What they are doing looks perfectly legitimate to me. Of course I could be missing something, but simply asserting (repeatedly) that there’s a “fallacy of assuming that what’s true for the individual case is also true in the aggregate” in these papers does not make it so.
18. November 2011 at 16:05
If getting Obama out of office is proof the tea partiers count then they don’t count as he will likely have a second term.
The “Silent Majority” idea has been around since Nixon. It hasn’t bared any fruit yet. Your assumption that you and your tea partiers count and no one else does is your illusion. Ultimately most Americans have a better view of OWS than the Tea Party. The Tea Party was nothing authentic it was bankrolled by the Kock Brothers and Alex.
There’s nothing grass roots about it was made up by somem oligarch’s marketing department. There’s a sucker born every minute so some like yourself think its real, but in truth it’s yesterday’s news. Obama is in good shape, the economy is improving, most Americans have a better view of Obama than either tea partiers or the rank and file Republican party. And like it or not, one person one vote regardless of bank account. In 2016 Obama will leave.
18. November 2011 at 16:06
Kevin Donoghue,
And the economy is made up of individuals. What is virtuous in an individual (cutting expenses in hard times and drawing back investment expenditure) can hardly be folly in the case of a country, right?
Equally, if you go from individual to individual, you find that successful theft enriches that individual. So, if you can compile a large list of people who have benefited from theft, does that prove that theft increases GDP?
18. November 2011 at 16:07
Kevin,
You haven’t responded to my comment. I give four reasons why those papers are suspect, one of which is the partial equilibrium/general equilibrium distinction (which is what we’re talking about here, formally).
18. November 2011 at 16:09
OhMy,
As I understand your debating strategy, it is “Assert what I believe. Then tell my opponent that they know nothing.”, sometimes alternated with “Tell my opponent that they know nothing. Then assert what I believe.”
18. November 2011 at 16:20
I’ll stick with my earlier tard analysis.
18. November 2011 at 16:35
Well none of your analyses make much sense but that one also shows your lack of class so go with it. Honestly discussing things with you is tiresome. yOu have nothing even remotely interesting to say.
If I needed this I would watch Sean Hannity. All the other discusssions people are having right now is more interesting than this one. Don’t bother me again or I will have to get a restratining order against you.
I want to discuss monetary policy not debate a poor man’s Rush Limbaugh
18. November 2011 at 17:26
[…] In the comments section of TheMoneyIllusion, Marcelo has this link to a Monetary Policy game. Marcelo says: Scott, Ezra Klein posted on twitter […]
18. November 2011 at 17:26
“And why not? The nation is made up of regions. What they are doing looks perfectly legitimate to me.”
All you’ve done is prove you don’t understand what “fallacy of composition” means.
I’d add that you don’t seem to understand why a simple regression of output/employment on stimulus funds (no matter how many covariates you add, which I don’t think Feyrer/Sacerdote tried to do, since that’d be not worth the effort) does not produce a necessarily useful multiplier. At the national level, it’s still arguable, which is why papers have been published on this, but you’re still contending with all sorts of confounding factors going on in the global economy, which we simply can’t account for. We try to ignore them, and ignoring them is arguable for an economy the size of the US, but it’s still difficult to take these numbers at face value.
And at the subnational level, it’s completely implausible to develop a national-level multiplier from regressions on regional-level data. It’s ridiculous to argue that because you can completely account for the effect of government spending in one region that you can scale that up to an aggregate level for the nation. You haven’t taken into account at all any effect the government spending might have outside the region it was ostensibly targeted at. When all these subnational divisions share a language, culture, identity, currency — and have a free flow of capital, labour and goods to boot — you have not captured the full impact of government spending by simply looking at how it changed things in one particular area.
18. November 2011 at 18:13
@Kevin Donoghue
I think you need to take a step back and think about the fallacy of composition a little longer.
Suppose there is an economy consisting of two states: California and Texas. California has 100 jobs making $1000 dollars worth of butter. Texas has 100 jobs making $1000 worth of guns. The government gives Texas $100 extra, and they end up with 115 jobs. California still has 100 jobs.
The stimulus had a multiplier of 1.5 in Texas. $100 created 15 more jobs. The fallacy of composition is to assert that California also would have had 15 more jobs, if both states had received $100. Of course the stimulus helped Texas…but it the lack of growth in California may also have been caused by the redirection of money to Texas. So you cannot say that the effects of stimulus on Texas would have been true for the combined economy of Texas and California.
This is exactly the fallacy committed by the papers using cross-sectional comparisons.
18. November 2011 at 18:46
On measuring inflation:
http://ftalphaville.ft.com/blog/2011/11/18/753971/on-misunderstanding-qe-and-uk-inflation/
Inflation is a non-thing – that cannot be measured.
18. November 2011 at 19:22
David, You said;
“You said leverage amplifies the effect of tight money; ergo, with leverage, any degree of (passive) tightening manifests itself as a larger decline in nominal spending.”
I was talking about regional effects. I said very specifically I don’t think they apply in the aggregate. The borrowers will spend less, but the lenders will spend more.
Kevin, You are wrong that at full employment there is no regional effects from fiscal stimulus. It would still occur. You’d have migration–or unemployment would fall below the natural rate.
They assume that if you spend more money in one area, it boosts GDP in that area. They infer that this means that if you spend more money overall, you get growth in aggregate. That’s what the multiplier is. It’s the aggregate effect. You keep saying there is no fallacy of composition, but they clearly claim the regional effect tells us something about the aggregate multiplier. That’s a textbook definition of the fallacy of composition.
Fed Up, Tight money is a policy setting expected to lead to below target NGDP growth.
Mike Sax, Morgan talks that way about everyone–including me. You need to laugh it off. I do.
Thanks Filipe, I’ll tell Mike you said hi.
Thanks Marcelo, That’s probably worth a post.
JimP, That’s right
18. November 2011 at 21:09
“The borrowers will spend less, but the lenders will spend more.”
If that were true, then leverage would have no impact on a decline on aggregate nominal spending. Since we know that leverage had an impact in certain geographies, and we know we had an overall decline in nominal spending, then it follows that lenders did not, indeed, spend more. Therefore, leverage did impact aggregate nominal spending.
18. November 2011 at 21:35
@Mike Sax
Your supposition about the Wall St. crowd being willing to cough up enough for a small tax increase doesn’t really answer the question of what good would that do for the OWS crowd. With the current state of financial affairs, ~$15T in public debt and ~$65T in unfunded liabilities over the next 20 years, it is pretty safe to assume that any new revenue has already been spent; a few billion in extra revenue will make very little difference in the face of a European style debt crisis happening here. Any new kinds of transfers would be simply taking from one transfer program to give to another, rearranging the deck chairs on the Titanic.
I’m sorry that the mob wants a warm fuzzy of spreading the misery (as if not nearly everyone took a huge hit in the housing/financial debacle – I am now surviving on fumes myself)instead of real solutions to our fiscal/economic problems, and that they really have no intention to fix anything. I think we have a serious problem when after their ideological counterparts in politics get routed in an election, they believe they should override it with obstruction of commerce and violence. No, that isn’t “what democracy looks like.” It’s what mobs looks like. It is the breakdown in the order of society and the rule of law which is not acceptable under any circumstances and should not be rewarded with any concessions whatsoever. If they want changes, they need to do it the right way, like the Tea Party did. If they can’t handle that and want to press the issue, there is only one ultimate outcome that people who do believe in democracy and order will accept. And no, that does not mean more goodies from the treasury.
18. November 2011 at 23:30
Scott,
I read these things from Krugman and I don’t understand how he can get them wrong. Even I know better, as someone who took only 2 econ courses way back in what seems like a lifetime ago. These freshman errors are even worse than his China bashing.
19. November 2011 at 00:31
Bonnie,
I’m unsure why you feel that way about the OWS crowd. The first two months were those kids playing dress up and having a really long sleepover. Sure, they dirtied up common areas, but in any sort of grand scheme, it’s something easily ignored. Indeed, commerce went on without issue. All of the violence we are seeing now was instigated by the police in a display of terrible executive judgment.
Why is it Americans celebrate the youth of Arab Spring, but pepper spray their own?
19. November 2011 at 01:42
@Scott: You keep saying there is no fallacy of composition, but they clearly claim the regional effect tells us something about the aggregate multiplier. That’s a textbook definition of the fallacy of composition.
You appear to have convinced yourself that all aggregation involves ‘the’ fallacy of composition. You say you get tired of correcting Keynesians who don’t understand how to estimate the multiplier. I suspect they get a little tired of your habit of inventing rules that they should be following.
@Integral: You haven’t responded to my comment. I give four reasons why those papers are suspect, one of which is the partial equilibrium/general equilibrium distinction (which is what we’re talking about here, formally).
Do you think that any of your for reasons supports Scott’s astonishing claim? He claims that the authors are “assuming that what’s true for the individual case is also true in the aggregate” and they are thereby falling into a well-known fallacy, which they undoubtedly learned about many years ago. There may indeed be problems with these papers but I see no reason to suppose that’s one of them.
19. November 2011 at 01:44
s/b “any of your 4 reasons”
19. November 2011 at 02:19
Brad DeLong with link to Daniel Kuehn; sensible responses:
http://delong.typepad.com/sdj/2011/11/estimating-multipliers-sub-national-cross-sectional-studies-are-a-perfectly-good-way-to-produce-a-lower-bound-to-the-constan.html
19. November 2011 at 02:57
Bonnie in your statement:
“Your supposition about the Wall St. crowd being willing to cough up enough for a small tax increase doesn’t really answer the question of what good would that do for the OWS crowd. With the current state of financial affairs, ~$15T in public debt and ~$65T in unfunded liabilities over the next 20 years, it is pretty safe to assume that any new revenue has already been spent; a few billion in extra revenue will make very little difference in the face of a European style debt crisis happening here. Any new kinds of transfers would be simply taking from one transfer program to give to another, rearranging the deck chairs on the Titanic.
You mistate the real problems facing the country. It is not debt and no we do not have a European style debt crisis here. If you notice our interest rates have not gone up as the deficit hawks(I notice they are only hawks during Democratic Administrations during Republican Administrations they declare “Reagan showed deficits don’t matter”) have warned with their Cassadnra scenario.
Indeed the worse things get in Europe the better our debt rating gets. If you noticed S&P’s downgrade did not make our treasury yields jump as they did after downgrades in Europe.
Right now would actually be a time for us to borrow more-creditors are effectively willing to pay us for our debt, it’s as if a credit card company says it will pay you about 5 percent per month on your outstanding balance.
There is not debt crisis. The scare talk then that our deficits will drive up the interest rates on U.S. Treasuries is not happening. The opposite is the case.
If you look at Europe, the ideology that tries to claim that what is happening there is all about out of control wellfare states in the EU is also wrong.
What we are seeing in fact is that not only countries with budget deficits and high debt ratios but countries that are very credit worthy like France, now or even Austria-Hayek country. It should be noted that countries like Ireland and Spain had budget surpluses in 2008 before the financial crisis hit. In truth a Euro country is hit regardless of whether it has a deficit or surplus and whatever its debt ratio.
This is why the austerity that they want to impose on Europe is so self-defeating. With Italian bond yileds as high as they are, no austerity package no matter how bloody will solve it. Their bond yileds have to go down it’s that smple.
Here in the U.S. there is no debt crisis. So your claim that raising taxes on the rich would be ” rearranging the deck chairs on the Titanic” is wrong.
19. November 2011 at 03:05
Scott you’re probably right. You yourself did however quote Keynes who said when talking to fools sometimes your words shuold be a little wild.
19. November 2011 at 03:13
By the way, Bonnie, the British too were stupid in imposing Euro style austerity on themselves when they have their own currency, printing press, and central bank.
I spoke of this more fully here.
http://diaryofarepublicanhater.blogspot.com/2011/11/britain-not-that-smart.html
On the differnce it makes-it makes all the difference in the world when a country like Britain maintain’s its own monetary frameworw-what Krugman callss the Rubcion Effect-please see here
http://diaryofarepublicanhater.blogspot.com/2011/11/eurozone-making-britain-look-smart.html
What changed between this last post and the one I wrote the next day above, is that in the first one I said the Euro crisis makes the British look smart for never getting into the Euro.
What I later realized though is that they makes them just the same doubly stupid for imposing “strcutural reforms” when they suffer from no debt crisis-there is no run on British debt.
19. November 2011 at 04:21
ChacoKevy,
One major difference between the Arab Spring and OWS is that the Arab Spring took place in countries with no peaceful routes of change. Disrupting Tahrir Square meant something quite different from disrupting Wall Street.
Now, it’s true that much of OWS is perfectly legitimate and positive political protest. It’s also true that there may have been considerable overreactions from the authorities in particular areas. However, I’ve not yet heard a persuasive argument for equivalence between the two protests.
Compare all this with the Tea Party: they have played the existing democratic system via organising voters and vetting candidates. You can agree or disagree with their goals, yet you can’t help but admire their methods. The Tea Party is real democracy in action.
It is interesting that, after all that (now defunct) “moderate our discourse” stuff at the beginning of the year, the American left now has its own rhetorically firery grassroots movement.
19. November 2011 at 06:02
Bonnnie,
It sounds as if you focus on macroeconomic issues for one of the same reasons as myself: total necessity. I didn’t begin these studies in earnest until every other life option had basically been exhausted. Not to say I don’t enjoy the mental challenges, but had always deemed the work of the mind a luxury given the need to survive.
In that spirit I’ll try to focus: what do we want most in the present? Is it not to avoid the possibility of falling back into the Malthusian trap that defined so much of human existence? I would dare to venture that – in order for us to remain in a state of continual economic and social advancement, women must remain part of the equation in real terms. That means remaining free to choose not just in monetary activity but non-monetary as well. You know all too well that both sexes, when they lose access in the first, they also lose access in the latter.
All political strategies aside, the Tea Party and OWS are two sides of the same coin of the cultural vacuum that presently exists. One says, how do we move forward with traditional values and a strong work ethic? The other says, how do we retain vital intellectual discourse as a part of our work ethic into the future? Clearly, we need both. NGDP targeting is a way to preserve the present. Understanding the agendas of Tea Party and OWS are ways to move into the future.
19. November 2011 at 06:06
ChacoKevy,
OWS is a joke.
Look, they could do like the Tea Party, but hippies without money protesting for free college is just dumb. The free lunch is OVER.
Repeat after me: free lunch is over.
That’s why the only meaningful claim the OWS makes, is where they agree with the Tea Party – no corporate cronyism, no gvt. bailouts.
OWS isn’t going to get a handout, but they can AT LEAST make sure someone else doesn’t get one too.
Now if OWS rose up and demanded that college debts be paid for the STEM graduates and going forward ONLY STEM graduates would get GVT backed degrees, that would be interesting….
19. November 2011 at 06:09
Becky,
“The other says, how do we retain vital intellectual discourse as a part of our work ethic into the future?”
I’m sorry but we’re not going to back up college degrees that don’t have marketable skills in a STEM based future.
Don’t put yourself on the side of insisting that kids get to study whatever they want and the economy fails if their skills aren’t worth anything.
19. November 2011 at 06:40
Morgan,
You just opened the door for what I wanted to explain to you! Unfortunately right now I’m going to give in to the desire to do so without the rumination of that last post but believe me I’ll hit you with rumination at some point.
Remember that lateral skills exchange economy I told you about? It is multi-tiered in ways that the top levels continue to be monetary as they are now, in that ego and curiosity are the perpetual motion machines of economic activity. However, for that machine to work in the aggregate, the playing field is opened up locally at non-monetary levels.
What’s special about those non-monetary levels is that they take place in ways that continually improve their efficacy over time (ego and curiosity need zero money to thrive). By so doing, wealth is always and exponentially increased. And through such a strategy, Earth need no more be the Poor Little Starter World that God and Jesus need to be perpetually sorry for.
19. November 2011 at 06:53
would instead prefer regulation of leverage, if FDIC and TBTF creates a moral hazard market failure.
leverage IS regulated, pretty heavily. Leverage is not the issue. whether you have one big bank, or 5000 small banks, the issue is that a significant segment of the economy crashed (residential real estate). When real estate crashed in the 80s, we had 800 S&L’s close up. The Fed (looking for more disinflation) was all too happy to allow the ensuing credit crunch. maybe its less noticible with 5000 small banks, but the effect is the same: if a large chunk of the economy goes south, banks will go south. the economy as a whole cannot diversify away that risk. in aggregate, someone is holding all those loans.
Now, suppose you are a regulator. Suppose you say “hold capital for the worst HPI downturn in 50 years.” HPI looked “terrible” through the 90s recession. even so, the risk on a loan looks very, very different in 2006 than 2011.
in short, regulating leverage (which we already do) is only as good as your ability to predict the severity of the next crisis. regulating lending standards (which also, the fed already does), is also only as good as your ability to predict default and recovery. now matter what leverage ratio you pick, its wrong if the crisis is worse than you think.
19. November 2011 at 07:04
Morgan,
Yes taxation is built in, and it also becomes direct democracy. One in five of our skill sets is provided to others based on our individual choice. Often that skill is chosen in terms of our current challenge, for instance, we might provide a skill in ‘tax’ terms that we are presently learning. If that is a health related skill, remember that the doctor is not always called upon to care for the patient at the top of her skill level, in fact she can tell you that often does not happen.
19. November 2011 at 07:10
Scott,
John Taylor criticizes NGDP Level targeting:
http://johnbtaylorsblog.blogspot.com/2011/11/more-on-nominal-gdp-targeting.html
Response?
19. November 2011 at 07:21
Morgan when you see me yawn when your talking that shows I’m interested.
Bonnie I appreciate that while I don’t agree with you put out a cogent arguement for “fiscal discipline” AKA austerity. However as I argued above, you’re wrong as the deficit hawks-and the inflation hawks-are like the French Magiinot Line or the Y2K bug: fighting an illness which doesn’t exist.
Your mistake is claiming that we are facing a European style debt crisis. No we are not as the U.S. Treasury yields show very clearly. Investors bascialy are paying us interest to hold onto our debt.
So I apreciate your coment as it has helped me to clarify much better the crux of the problem and why exactly the conservative Republican alarmism about deficits and the need for “entitlement reform’ is wholly mistaken.
I have written a post where I have expanded and clarified thuis point.
http://diaryofarepublicanhater.blogspot.com/2011/11/no-us-has-no-euro-style-debt-crisis.html
You have however helped me locate the exact point of error of the tea “structural reform” rap.
19. November 2011 at 08:02
“OWS is a joke.
Look, they could do like the Tea Party, but hippies without money protesting for free college is just dumb. The free lunch is OVER.
Repeat after me: free lunch is over.”
I agree, and I think the college example is especially sharp. College-grade education is about to become a near-free lunch thanks to efforts like khanacademy.org (started by a hedge fund quant, so OWS gets no credit here).
Certification and credentialing will always involve some cost, and the government could play a useful role here by supporting smaller and leaner certification agencies. But there is no case for supporting traditional colleges which will always charge inflated admission prices (due to their enduring prestige, protecting them from competitors with a better cost structure). Yet this is what the “free college education” protests call for.
And this sort of economic irrationality is pervasive in the OWS protests. They don’t care about free lunches even when they actually exist and are easily found; all they want is “FREE MONEY and “FREE $#!+” paid for by the taxpayer.
19. November 2011 at 08:21
David, You said;
“If that were true, then leverage would have no impact on a decline on aggregate nominal spending. Since we know that leverage had an impact in certain geographies, and we know we had an overall decline in nominal spending, then it follows that lenders did not, indeed, spend more. Therefore, leverage did impact aggregate nominal spending.”
More fallacy of composition. In the localities where NGDP was affected the borrowers and lenders didn’t balance out. That’s the whole point of my comment. Nevadans are net borrowers, Massachusetts residents are net lenders.
Mike, Sometimes really smart people are so convinced they are right that they don’t look at things from the perspective of the side they disagree with. If you simply asked Krugman how Barro would respond to that study, if you asked him to put himself in Barro’s shoes, then Krugman would see the fallacy of composition in 10 seconds.
Kevin, I have a very simple question for you. Imagine the GDP of Podunk county is $100,000,000. Now assume that the government builds a $5 billion military base in Podunk, and assume the money comes from non-Podunk residents. Please tell me precisely which anti-Keynesian model denies that the GDP of Podunk would rise. Once you show me the specific model that Krugman and DeLong are criticizing that would make that nutty prediction, maybe we can advance the debate. Maybe we can see whether it really does imply that Podunk’s GDP should not rise.
DeLong’s made that mistake before, he simply ignores the criticism by me, Ambrosini, and even the Kuehn comment to his post.
These Keynesian studies simply assume the multiplier was positive, as if it was zero YOU WOULD STILL GET POSITIVE RESULTS AT THE LOCAL LEVEL.
Mike Sax, Fair enough.
dwb, You said;
“leverage IS regulated, pretty heavily. Leverage is not the issue. whether you have one big bank, or 5000 small banks, the issue is that a significant segment of the economy crashed (residential real estate). When real estate crashed in the 80s, we had 800 S&L’s close up. The Fed (looking for more disinflation) was all too happy to allow the ensuing credit crunch. maybe its less noticible with 5000 small banks, but the effect is the same: if a large chunk of the economy goes south, banks will go south. the economy as a whole cannot diversify away that risk. in aggregate, someone is holding all those loans.”
This is exactly right, and indeed I’ve made the same argument. The sort of leverage that needs to be restricted in subprime mortgages. I’ve suggested banning any mortgage with less than 20% downpayment unless it uses non-FDIC-insured funds. But “derivatives” aren’t the main problem, they just move risk around. So I agree with you there.
Thanks Eric, I did a post on that this morning.
19. November 2011 at 10:18
Scott,
Let’s see:
-borrower states experienced bigger declines in spending
-lender states did not make up for that spending
-therefore spending declined nationally
Arguing the irrelevance of debt assumes a frictionless economy. The obvious reality is that financial intermediation shrank as the shadow banking system largely disappeared.
19. November 2011 at 10:52
Scott: Please tell me precisely which anti-Keynesian model denies that the GDP of Podunk would rise.
None that I know of. You evidently think I suppose otherwise.
19. November 2011 at 12:14
Scott,
After toying around a bit more with the game, I got an ‘oil shock’ scenario, and the game’s appropriate policy was to increase rates to ‘stop inflation’. This is a pretty clear case of ‘supply-push’ inflation, and the game actively promotes poor policy responses to these scenarios. I know that the people in the actual monetary policy committee must know more than that, but still I think that this shows a total lack understanding even a fundamental AD-AS type model…
Also, if this one seems bad, I read on Matt Yglesias’ blog about an ECB app in which the policy incentives were even more perverse (http://thinkprogress.org/yglesias/2011/11/18/372102/economia-the-game-of-terrible-monetary-policy-lessons/) he does a good job of showing how bad the lessons from the game are.
Thanks!
19. November 2011 at 12:54
Marcelo,
I took a ‘do nothing’ approach (kept rates the same) to the oil shock and eventually it worked itself out and prices came down.
Fed guru was reappointed.
19. November 2011 at 13:08
Dustin,
I think that my point still stands that if the Central bank were to tighten during an oil crisis (if there was a negative shock to NGDP) then that would lead to a deepening of the recession, but the game gets this concept completely wrong. I imagine that when the oil shock happened the game saw unemployment fall, not rise, definitely not what would happen in reality!
Interesting that the ‘do nothing’ approach is so successful (still probably saying something about how the Fed acts today!). Also, the Fed guru being appointed, despite not doing anything to use the Fed’s ability to minimize the cost to society welfare from the oil shock also speaks to how central banks see themselves and their responsibility towards society.
19. November 2011 at 18:35
David, Spending fell nationwide because of tight money–debt had nothing to do with it. Debt tells us why some people cut back while others increase. If the total declines it’s not because of debt, it’s because the Fed isn’t doing NGDP targeting, level targeting.
Kevin, If even models that predict a zero aggregate multiplier also predict local multiplier effects, then finding local multiplier effects tells us what?
Marcelo, Yes, I also saw Yglesias’s comment on the ECB game. I’d tried playing but couldn’t get interested.
20. November 2011 at 02:41
Scott, a model that predicts a zero aggregate multiplier necessarily implies that if one region’s multiplier is positive, some other region’s multiplier is negative.
20. November 2011 at 02:47
To clarify that: of course I’m not saying that the aggregate multiplier is a simple sum of regional multipliers; I’m merely saying that you can’t get a zero aggregate multiplier if any regional multiplier is strictly positive and none is negative.
20. November 2011 at 04:39
@David Pearson:
“The more stable NGDP growth, the higher the leverage, the harder it is for the Fed to hit an NGDP target without significant overshoots on both sides.
Ergo: the problem is debt.”
Agreed!! I’ve been arguing on this blog that debt stock is fundamental.
Higher leverage makes one more susceptible to shocks. If NGDP succeeds in smoothing out such shocks, then all that happens is you trade short term smoothness for more systemic and tail risk.
“Arguing the irrelevance of debt assumes a frictionless economy. The obvious reality is that financial intermediation shrank as the shadow banking system largely disappeared.”
I would go beyond that and say that even in a frictionless economy debt stock is absolutely fundmental.
20. November 2011 at 04:46
[…] Scott Sumner has been very critical of the studies (many of which are favorably reviewed by Christy Romer) that purport to show the […]
20. November 2011 at 06:30
Doc Merlin,
I completely agree with what you say about debt stock, it really has become my prime way of understanding NGDP. Which also means income cannot be lowered too much without losing the ability to repay debt. That is also why income can not go too far down the road of ‘watering down’ by redistribution (lowering production efficiency) or liquidation by auction. While wage stickiness puts more people in a bind in an immediate sense (auction liquidation or redistribution of income can be temporary solutions) the greater solution is to tie more assets to pay as you go methods in the future, so that sticky wages do not become a problem in and of themselves.
20. November 2011 at 07:24
Kevin, Agreed, but how does that relate to these cross-sectional studies? They’d have to take a stand on the aggregate multiplier, and then we are right back where we started, with all the difficulties of measuring aggregate multipliers.
20. November 2011 at 08:31
Scott, I don’t see why such studies have to “take a stand on the aggregate multiplier” if what you mean by that is pinning it down to a specific value. For example, suppose a UK study finds regional multipliers of 0.5, 0.7, 0.9, and 1.1 for England, Wales, Scotland and Northern Ireland. That doesn’t tell me what the UK aggregate multiplier is, but it surely implies that it is positive.
20. November 2011 at 10:31
“That doesn’t tell me what the UK aggregate multiplier is, but it surely implies that it is positive.”
No, it doesn’t. You recognise that you can’t sum up/average the regional multipliers to get a national multiplier. The reason for this is that you are getting the coefficients for the marginal impact of stimulus spending in Wales on the economy of Wales, marginal impact of stimulus spending in Scotland on the economy of Scotland, etc. You are not getting the coefficient for the marginal impact of stimulus spending in Scotland on the economies of England, Wales and Northern Ireland.
Regional-level studies do not tell us anything about the national-level multiplier.
20. November 2011 at 11:15
I’ve suggested banning any mortgage with less than 20% downpayment unless it uses non-FDIC-insured funds.
I do not think that would have prevented the subprime crisis for several reasons:
1. many of the subprime loans were based on fraud (“liar loans”). Income, appraisals, assets, were all fudged. The fact is, the GSE’s, the Fed, and the OCC were not pushing for enough due diligence and documentation, and if they had they would have discovered a lot of early rot. If I trade-up my house, and fudge the appraisal, it can look as though I have 20% down, no problem. If HPI goes up 20% then when i trade up the loan to value looks great as well.
2. For GSE mortgage pools, they ARE reinsured so that the GSEs do not take the first 20% loss. That is, if the loan is 95, the house is worth 100 (95% LTV) then the next 15 is reinsured (through PMI or some other company). PMI Corp just went into receivership, of course. For private label securities, they can also be reinsured, or sliced and diced so that the investor buys a tranche that continues to pay out even if (for example) 50% of the loans default. In theory, the purchase of reinsurance or tranching allows the bank to claim that FDIC funds are not at risk for the top 20% of the losses (which is true – someone else bears the risk on the fist 20% of losses).
3. banks cannot really directly tie deposits to loans. There is a big pool of assets and liabilities, but one cannot say that these x loans were made with these y liabilties (deposits).
4. I’ve heard anecdotal evidence banks were undercharging for the loans, supressing credit officers views that these loans were risky and undercharging for the risk to get in the market.
Also, its not just defaults, but recovery rates that matter too. An 80% LTV in flordia on a condo can be more risky than a 95% LTV in the midwest on a single-family property.
A rule banning any mortgage with less that 20% downpayment unless it uses “non FDIC funds” is equivalent to a rule that requires banks to hold more capital for riskier loans. Again, we have such a rule. Having a bunch of hedge funds issue subprime mortgages doesn’t help either, when they crash it can be equally destabilizing. Should we just not originate loans unless there’s a 20% downpayment? If I have 1 million dollars in the bank, and I want to buy a 500k house and borrow 500k to do it, and I can find someone willing to take the first 100k in losses if i don’t pay, i cannot get a loan? A credit officer would properly say I need to have skin in the game (reputation or otherwise). But – I don’t think the answer is no- the interest rate I get charged should reflect the risk.
If you add up the amount of fraud and deceit that was going into originating these loans, I do not think such a rule would be effective. I think having consistent forward-looking lending standards, charging properly for risk (including tail risk and the risk that the HPI forecast is wrong!), and performing adequate upfront due diligence solves the problem. more regulations will only constrain the law-abiding. the portion of the industry that was willing lie will find a different clever scheme to flout the rules. thats why in my mind the focus should be on enforcing the rules we have and making sure that due diligence and audits are far more rigorous.
21. November 2011 at 06:57
@johnleemk, your comment offers a clue as to what Scott may be trying to say, so thanks for that. I’ll seek clarification in his new thread, where he adds Christy Romer to his list of people who fall for this alleged fallacy of composition.
21. November 2011 at 13:51
Kevin, See Johnleemk’s comment.
dwb, I don’t agree. There weren’t a lot of sub-prime loans back when I got my mortgage in 1991, because you had to put at least 20% down to avoid mortgage insurance. When that requirement was dropped sub-prime loans became much more popular.
The failure of the PMI company supports my argument, as the alternative proposal would be to go back to forcing sub-20% down loans to have mortgage insurance. But if the insurance companies are going to fail when you most need them, better to ban sub-prime loans entirely.
I don’t claim that this would solve the entire problem. I’d also like to ban banks from using FDIC-insured accounts to make commercial loans. (They could still use funds raised in other ways.) Many people don’t realize that commercial loans, not sub-prime loans, were the biggest problem, the biggest source of bank failures.
I have no problem with your suggestion that we enforce rules against fraud, and throw the bums in jail if they get caught.
That info on the GSEs is very useful; it certainly suggests they purchased a lot of toxic mortgages, which goes against the prevailing mythology on the left.
21. November 2011 at 18:13
[…] Scott Sumner responds to the study I mentioned yesterday, none too favorably: And let me guess, counties with economies dominated by autos and steel usually suffer bigger job losses in recessions than counties dominated by hospitals and colleges. Does that tell us anything about what causes recessions? Doesn’t EC101 also teach that correlation doesn’t prove causation? […]
21. November 2011 at 20:44
All GSE conforming loans (fannie, freddie, ginnie) loans have required mortgage insurance when the LTV was higher than 80%… that never changed. FHA has had a sub 5% downpayment program for most of its history. Even so, PMI costs (say) $115 per 1000 borrrowed. there is really no difference bewteen that and paying 1.15% more in interest.
LTV is only one dimension of the credit decision. there is also your credit score, income, and assets. many sumprime loans actually had decent LTVs (50%-60%) but bad credit (generally <600 FICO score). low LTV does no good when you lose your job and the housing market plummets 33% (60% in some places).
there are papers from the era you can pull from the wayback machine (I served time in the industry). table 4 from this paper shows that 42% subprime were low/nodoc with a sub-620 credit score. omg! A lot of lenders were allowing "low doc" and poor credit with low LTV ratios (bigger downpayment, see the st louis fed paper).
Now, if someone wants to buy a house with 5% down and good credit, why not? whether they get charged $200 per 1000 in PMI or 2% more in interest rate, its the same. If someone has a 500 credit score, thats ok too but I would charge them 15% not 2%. Proper risk-based pricing chokes off demand for these loans.
the issue is: 1) the industry was systematically undercharging for the risk due to rosy HPI assumptions and 2) low-doc became synonymous with fraud (how hard it is to provide last years tax return or some proof you have income?). and if you dont believe the income, you should not believe the 65% LTV either.
if the industry had properly charged for the risk, the demand would have choked itself off. if you want to talk about leftists, the leftists were happy to have people get into mortgages and when you said i need to charge grandma 15% over a prime mortgage, there were cries of predatory lending.these are the same people who compain that they should have not been allowed to get a mortgage because the forclosure ruined their credit.
The GSE's did not do much subprime because they were embroiled in their own morass of accounting issues, until way too late in the game to have an impact.
and all those prime mortgages … well 60% LTV does not do any good if you lose your job and HPI goes down 50% in your area (say, FL). even with 80% LTV suddenly we are all subprime!
http://www.finance-innovation.org/risk09/work/2504430.pdf
http://research.stlouisfed.org/publications/review/06/01/ChomPennCross.pdf
22. November 2011 at 07:21
dwb, You said;
“Now, if someone wants to buy a house with 5% down and good credit, why not? whether they get charged $200 per 1000 in PMI or 2% more in interest rate, its the same.”
It may be that mortgage insurance is the same as a high rate (but that assumes the mortgage insurance companies would always fail in a crisis–if this is not the case then mortgage insurance reduces risk of bank failure.)
But more importantly I am suggesting a ban on those loans. Even with higher rates or mortgage insurance, those loans are risky, and lead to a moral hazard problem (taxpayer bailouts of failed banks.) So ban them entirely, at least for institutions using FDIC insured funds. I’m happy to allow those loans from non-FDIC insured, non TBTF, institutions.
I agree that 20% doesn’t solve the problem, you still need stable NGDP growth, good monetary policy, so that home prices have less wild swings up and down. And as you said you need tighter regulation against mortgage fraud. Fraud is something that even libertarians favor regulating against.