They run a brilliant blog. Too bad it’s marred by unsubstantiated daily rants against Keynesianism…
Let’s make no mistake about this, the ZeroHedge blog run by Tyler Durden (or several Durden’s, as it happens) is brilliant. But it is marred by daily rants against Keynesianism. That wouldn’t be so bad if these rants were actually substantiated by facts and arguments, but we couldn’t really detect much, if any. That Keynesianism doesn’t work seems to be assumed to be self-evident at ZeroHedge.
We even tried to engage these Tylers into a discussion about the topic, but they didn’t take the bait seriously. So we’ll debate from here.
The rants against Keynesianism are so frequent, we don’t really have to provide many examples. Just one will do, to give you an impression:
- And so, as any remaining voices of reason realize they are dealing with a group of deranged Keynesians, soon there will be nobody left in the administration who dares to oppose the destructive course upon which this country has so resolutely embarked, which ends in one of two ways: debt repudiation, or war. [ZeroHedge]
Now, one of those ‘deranged Keynesians’ is Larry Summers, no less. Not exactly a Keynesian (or any other kind of) radical, and indisputably one of the countries most distinguished economists. We say this because ZeroHedge (albeit not Durden himself) did have no qualms in undressing another economist academic achievements, those of James Galbraith.
This is, of course, a bit harder to do with Larry Summers or Paul Krugman.
Here a typical commentary (albeit not from ZeroHedge) for those holding this view:
- I’ve insisted all along that the US should have allowed the primary bear forces to fully express themselves, as they inevitably will do anyway. But in its arrogance and ignorance, the administration decided that they could halt or sidestep a recession by printing us out of trouble. It’s been a terrible and expensive mistake…. What’s next? I think Washington will continue trying to spend us out of recession. This did not work in the past, and it’s not going to work now.
First off, whether a particular economic policy isn’t working is extremely difficult to ascertain, all kinds of statistical heroics need to take place and multiple studies need to confirm as few of these ever become clear-cut. The problem with economics is that there are just too many variables, and many are interdependent.
Even if it’s not working, it might very well be because the stimulus was badly designed. For instance, tax cuts (at least 1/3 of the stimulus package) could very well have been mostly saved. Pretty plausible with an over-leveraged household sector in the face of the economic mayhem that was going on. Many Keynesians also argue the stimulus wasn’t large enough.
We haven’t heard many claiming that the stimulus didn’t work pretty well in China, were it was proportionally significantly larger.
A good place to start the assessment of whether the stimulus worked is:
- To come up with a diagnosis of the situation
- To compare similar situations and look at policy differences and outcomes
First, the diagnosis.
Proper diagnosis: balance sheet recessions
Before slashing the remedies, it might be worthwhile to actually devote some time figuring out what went wrong with the economy. We won’t really deal with the origins of the financial crisis, but one thing is not in dispute. Over-levering of households and financial institutions has played a major role.
It might therefore be worthwhile to read some economist describing ‘debt deflation’ like Irving Fisher in the 1930s, or Richard Koo‘s related work, like his “Balance Sheet Recession” or “The Holy Grail of Macroeconomics: Lessons from Japan’s Great Recession”. If one doesn’t have time for that, have a look at his presentation called “The Age of Balance Sheet Recessions: What Post-2008 U.S., Europe and China Can Learn from Japan 1990-2005”
More especially exhibit 6, where Koo demonstrates that despite huge private sector saving to pay down debt (even at zero interest rates), GDP kept on growing, and exhibit 7 shows how they pulled that off: massive government spending. Exhibit 10 shows how premature fiscal tightening in 1997 and 2001 worsened things.
The question is, of course, whether America is in a similar situation as Japan. Well, if not, things are eerily similar, and in some expects actually worse (Japan doesn’t depend on foreign capital).
The main cause of a balance sheet recession is that the private sector has bought assets with borrowed money, and something (usually a tightening of monetary policy) triggers a fall in these asset prices. This easily leads to a vicious cycle with worsening private sector balance sheets and de-levering, forced selling of assets, further asset price falls, further saving to repair balance sheets, which transfers to the real economy in lower spending and to the banking system in falling credit demand and a worsening of bank balance sheets.
It could get distinctly worse if, apart from falling asset prices by efforts to repair balance sheets, the resulting decrease in spending triggers the general price level to fall. That would actually increase the real value of the outstanding debt, a really serious risk best to be avoided.
Albeit not by monetary policy, as Koo has a further persuasive argument: monetary policy is ineffective under a balance sheet recession as credit demand falls, even at zero interest rates (see exhibit 19). Now doesn’t that sound like Keynes “pulling on a string?” Monetary policy does have some use by injecting capital into the banks to avoid a credit crunch (exhibit 12-14) and weakening the external value of the currency (exhibit 20).
We have to add here that Tyler Durden from ZeroHedge doesn’t seem to disagree with Koo’s analysis (see here, here, and here), yet violently disagrees with the policy remedy offered by Koo, which seems to follow rationally from the analysis. Tyler offers only a single argument (apart from mocking):
- Keynes’ ideas may have been an operable theory when the world was not leveraged 100% debt/GDP (and 400% total debt including assorted off balance sheet items). Now, it is not. And everyone who blindly pushes for endless stimuli will find out that the end-play to Keynes’ fatally flawed economic theory is sovereign default. [Tyler Durden]
This is odd indeed. Koo’s analysis is that the private sector is over-leveraged, embarking on de-levering, which results in a fall in demand (as well as lower credit demand and asset price falls). Rather than being an obstacle, the private sector de-levering is actually the rationale for the public sector to step in to fill the spending void…
What would have happened without expansionary policies?
Saying that Keynesian policies aren’t working (even if that can be conclusively shown) is not enough. One has to show there were better (non-Keynesian) alternatives available. Of course, we can never tell what would have happened if countries wouldn’t have embarked on Keynesian stimulus, but perhaps the 1930s is a good reminder as any.
There is a certain amount of superficial elegance in arguing that if the cause of most of the problems is over-borrowing, certainly we shouldn’t borrow even more, right? The immediate riposte against this is that sovereigns are usually able to borrow at substantially lower cost compared to the private sector.
More importantly, Koo, like the Keynesians, argues that the only rescue can come from increased public spending (not tax cuts, these will just be used to repair balance sheets). Koo argues that:
- “Had there been no fiscal stimulus,” surmises Koo, “the Japanese economy today would have contracted by 40-50%, if the U.S. experience during the 1930s is any guide.” [Japan Review]
He also shows that, contrary to many perceptions, the medicine has done a great deal in alleviating the problem in Japan as balance sheets have steadily improved (exhibit 8).
There are some differences, like in Japan the main borrowers were corporations, while in the US the over-leveraged parties where households (buying real estate) and banks (“de-risking” these and thereby over-levering). Firms are actually hoarding (a lot of) cash [see Pragmatic Capitalist]
But apart from that, the drivers in the form of de-levering, falling asset prices, falling credit demand despite near zero interest rates, falling demand (at least by the private sector), increasing banking problems, etc.
However, the situation is similar enough, and the risks described (the debt-deflationary vicious cycles) scary enough to take anyone raging about “deranged Keynesians” without providing any alternative (or even showing where the Keynesians are wrong) with a considerable amount of salt.
That 30s feeling..
Apart from the Japanese experience which we’ve just discussed, another balance sheet recession (or, rather, depression) was the 1930s. It has been shown that the fall of aggregates like industrial production, stock markets, and world trade are as steep if not steeper compared to the first period of the 1930s depression. 
- If anything, the initial stages of our own recent crisis were more severe than the Great Depression. Global trade, industrial production and stocks all dropped more in 2008-9 than in 1929-30, as a study by Barry Eichengreen and Kevin H. O’Rourke found.[New York Times 22]
Now, what might have stopped the rot last year? Isn’t the rather swift policy reaction one of the main, if not the main difference from the 1930s? Doesn’t that at least strongly suggest that Keynesian policies have prevented us from falling into the abyss that was the 1930s depression?
Isn’t it therefore wholly premature, to say the least, for those arguing that it hasn’t worked? But there is more.
The 1930s vicious cycle only started to break when policies were turned around, and although tight fiscal policy was by no means the only policy mistake early on (Ben Bernanke famously blamed the Fed for the depression, and adherence to the gold standard and rampant protectionism were other big mistakes), people who have looked at the data have concluded that:
- Where tried, fiscal policy was effective in the 1930s [Almunia et. al.]
By ‘fiscal policy’ they mean expansionary, Keynesian policy.
The alternative, supposedly is fiscal retrenchment, to avoid “leading us to debt repudiation or war,” according to the alarmist ZeroHedge article quoted above. The crucial issue is, of course, whether the private sector will take up the slack. If Koo is right and the private sector is too busy repairing balance sheets, this is extremely unlikely. In fact, all that public spending is there exactly because the private sector retrenches.
We couldn’t find a discussion on any of these issues on ZeroHedge, but even they can’t deny the following logic:
- “A reduction in the fiscal deficit must be offset by shifts in the private and foreign balances. If fiscal contraction is to be expansionary, net exports must increase and private spending must rise, or private savings [must] fall.[Martin Wolf in the FT]
This indeed is the central issue. How likely is it that the private sector will increase spending in the face of fiscal contraction (provided there is little chance foreign demand won’t take up the slack as supposedly everybody should be waned off the Keynesian delusions).
Some, like Goldman Sachs have come up with studies containing examples where fiscal contraction turned out to be expansionary. However, as Martin Wolf pointed out:
- Thus, experience of fiscal contraction is going to be very different when it occurs in a few small countries… when the financial sector is in good health… when the private sector is unindebted… when interest rates are high… when external demand is buoyant… and when real exchange rates depreciate sharply….” [Martin Wolf in the FT]
Here’s Krugman tearing these to shreds:
- So every one of these stories says that you can have fiscal contraction without depressing the economy IF the depressing effects are offset by huge moves into trade surplus and/or sharp declines in interest rates. Since the world as a whole can’t move into surplus, and since major economies already have very low interest rates, none of this is relevant to our current situation.”] [Krugman]
The effectiveness of public spending
There are, of course, perfectly reasonable economic arguments against public spending as a way to increase economic activity. They fall into three categories, although they all have in common that increase in public spending will somehow be offset by a decrease in private spending:
- Crowding out
- Ricardian equivalence
- Say’s Law
It comes in different version, the most basic one argues that resources used in public spending cannot be used in private spending, or drives up prices of resources and thereby reduces private spending. This makes little sense when there are ample resources lying idle.
The main crowding out mechanism argues that bond-financed public spending claims a greater part of the savings pool, pushing up real interest rates and thereby reducing private spending. Have we seen interest rates increasing?
In an age of free-flowing capital, this argument makes less sense, as savings can come from overseas. One could still argue that tapping overseas savings to finance public spending could rise the real exchange rate, and thereby ‘crowd-out’ private spending via the trade balance.
But have we seen an enormous rise in the dollar as a result of stimulus spending?
The US did, by the way, have large trade deficits even in the final years of the Clinton administration when the Federal budget was actually in surplus.
Crowding out arguments make (some) sense in times when the economy is purring close to full capacity and resources are constrained, but not in a depressed world with idle resources. That hasn’t stopped the austerity people invoking other arguments though.
This curious theorem comes from the wet dream of conservative economist creating a parallel universe in which all markets clear instantly and all people are completely rational agents so their savings match expected tax increases as a result of stimulus spending dollar for dollar.
Do you know anyone who started to save more because of the stimulus spending? Has the idea ever even occurred to you?
Yet a surprising amount of politicians seem to ascribe to some form of this theorem, says Merkel:
- Reducing the budget deficit by 10 billion euros ($12 billion) per year “won’t put a brake on the world’s economic growth,” Merkel said, relating what she told Obama yesterday. Germans are more likely to spend money if they feel the government “is taking precautions” to ensure solid finances, she said. 
The ‘confidence’ argument used here is just Ricardian equivalence in reverse.
We think Krugman has done enough to show that even if this curious theorem actually holds true, a temporary stimulus will still be a net stimulus:
- It’s one thing to have an argument about whether consumers are perfectly rational and have perfect access to the capital markets; it’s another to have the big advocates of all that perfection not understand the implications of their own model. [Krugman]
There are even those that argue that:
- debt-financed government spending necessarily crowds out an equal amount of private spending, even if the economy is depressed — and they claim this not as an empirical result, not as the prediction of some model, but as the ineluctable implication of an accounting identity. [Krugman]
You should read the rest of the article to appreciate the nonsense of that. The proponents mistake an identity for a behavioural relationship. Actual savings plus tax receipts indeed equal actual investment plus public spending, but that’s an identity. It says nothing about the behavioural relationship between the these variables.
So an increase in public spending could very well increase GDP, thereby increasing savings and tax receipts (so the identity holds), rather than necessarily reducing investment.
Simple budgetary arithmetic
It turns out the stimulus is not terribly important in the greater scheme of public finances:
- Consider the long-run budget implications for the United States of spending $1 trillion on stimulus at a time when the economy is suffering from severe unemployment. That sounds like a lot of money. But the US Treasury can currently issue long-term inflation-protected securities at an interest rate of 1.75%. So the long-term cost of servicing an extra trillion dollars of borrowing is $17.5 billion, or around 0.13 percent of GDP. And bear in mind that additional stimulus would lead to at least a somewhat stronger economy, and hence higher revenues. Almost surely, the true budget cost of $1 trillion in stimulus would be less than one-tenth of one percent of GDP. [Krugman]
- IMF assesses that from a total of 35.5% increase of public debt as a % of GDP in the years 2007-14, only 3.5% comes from stimulus. [Krugman]
The risk of deflation
- Let’s recap. Unemployment is high and is in reality going higher if you count those who would take a job if they could get one. Incomes are weak. Plans to purchase discretionary items are falling. Housing is likely in for a further drop in prices. The stock market is not exactly booming. Treasury yields are falling, not from a credit crisis or a flight to quality, but because of economic conditions (deflation). Money supply is flat or falling. Prices are under pressure. The list goes on, and all factors are indicative of deflation. [John Mauldin on Pragmatic Capitalist]
To embark on universal, hence synchronized fiscal austerity in this environment bears a considerable risk of plunging the world (at least the Western world) into a deflationary-debt spiral from which it’s really hard to escape. Public finances will not improve, but that will only be ‘collateral damage’ compared to the carnage done to the real economy if this happens.
The limits of Keynesianism
Yes, there are limits. It doesn’t mean the theory is defunct. Some of the PIIGS are facing these limits (in the way of meeting actual, rather than imaginary bond vigilantes), but this is mostly produced because:
- They can’t depreciate their currency (compare the UK, which is the master of it’s own monetary and exchange rate policy, with Spain, having comparable public finance problems but trapped inside the euro)
- They’ve lost a great deal of competitiveness vis-à-vis the core euro countries, so even if they could inflate their way out of the public finance mess, that would aggravate the real economy.
We don’t think there’s any Keynesian who would argue that, say, Greece would have to embark on massive stimulus spending. The point about Keynesianism is that there is another part of the equation, countries have to run surpluses during the good times in order to be able to afford to embark on. That many didn’t do that can’t be held against the Keynesian theory and places greater responsibility not only on them, but also on the countries that did.
It might also be useful to keep in mind that many countries, including the UK and the US, had worse public debt situations after WOII. What happened is that a combination of economic growth and moderate inflation steadily reduced the real value of that debt as a percentage of GDP.
The positive case for public spending
We wouldn’t be surprised if many of the ‘austerity now’ advocates are also of the persuasion that government is always the problem, never the solution. These people might actually be reminded of the positive contributions to economic growth that public investments could bring (and have brought in the past).
We could cite lower transaction cost, valuable infrastructure, addressing all kinds of market failures like collective action problems, external effects, or information asymmetries (a prime cause of the financial crisis, as we wrote before). We could cite the GI bill, the education system that the US build a century ago which was an important factor propelling its economy higher, the things coming out of DAPRA, etc. etc.
One could argue that just like Japanese firms have the habit of increasing maintenance and training during slack times, public investment should be increased during slack times, not only to make up for the spending shortfall, but to upgrade the economic structure and thereby laying the foundations for future economic growth.
The Chinese are doing it. (Here is only one example, the amount of funds going into laying the foundation for knowledge-intensive industries is quite impressive.)
And as long as the returns from these kind of public investments are larger than the interest rates on the debt used to finance these (and for education and R&D spending there are numerous indications this is the case), we will actually improve public finances over time.
Last, but certainly not least, one should have a look at the countries that did actually embark on austerity, like Ireland, Estonia, Latvia. These are small open economies, so their examples are best case scenarios as their austerity was at least partly offset by increasing economic conditions elsewhere. In fact, Leo Kolivakis just wrote this on ZeroHedge:
- Krugman is right about Ireland, Latvia and Estonia. They all implemented savage cuts, unemployment went up, as did the cost of insuring their debt, and government revenues dwindled. It has been nothing short of a monumental disaster.
Now wait until most of the developed world will embark on austerity, instead of just a few small open countries free riding on the stimulus of others. The result will be much worse. Which is perhaps why:
- the bond market is more worried about a 1930s echo right now, which is driving yields lower. If they were more worried of massive fiscal crisis leading to a run on the US dollar, then yields would be skyrocketing up, not down. [Kolivakis: The Myths of Austerity]
- There is little evidence embarking on austerity, instead of Keynesian demand management as a reaction to a private balance sheet recession provides better economic outcomes.
- If the 1930s or Japan 1990-2010 are any guides, the evidence points squarely in the other direction.
- The few austerity cases today are best-case scenario’s for the austerity camp (as they’re open economies and essentially free-riding on the stimulus of others) don’t provide much hope either.
- The austerity camp seems to be blind to the debt-deflationary vicious cycle risk of a universal austerity policy in the Western world.
- The numbers actually show that the discrete Keynesian spending is a small part of the public finance problem.
- The austerity camp also seems to be blind that public investment, apart from making up from decreased private spending, could generate positive longer-term returns for the economy and thereby public finances if the return on the public investment is larger than the interest paid on the debt issued in order to finance it. Considering the low interest rates and public investment categories like R&D and education which are generally recognized to have significant returns, this seems quite plausible.
- The theoretical underpinnings of the anti-Keynesian camp is pretty shaky to say the least. They have certainly not shown how public spending crowds out private spending in situations of underused resources, or that people will offset public spending to save for future tax increases, or that Say’s law holds.
- The examples of cases where austerity turned out to be expansionary depend on the private sector and/or overseas spending to compensate reduced public spending. Most of these cases involved either sharply lower interest rates and/or increased competitiveness. Interest rates are already very low and there are few currencies against which to devalue left if the whole Western world embarks on austerity, so these examples are not likely to be very relevant.