Whenever we write an article, we get a lot of rap for supposedly being Keynesians, even when the article in question doesn't contain anything particularly Keynesian. While we wouldn't really define ourselves as Keynesians, we do think Keynesian economics is the right answer for the present economic circumstances. Since we get a lot of criticism for that, it's about time for a larger expose on this.
We're all Keynesians now
This is what a Republican(!) U.S. president argued in 1971, and it wasn't even much of an exaggeration, if any. Here is a prominent Keynesian of today:
As recently as 2004, the Economic Report of the President (pdf) of a Republican administration could espouse a strongly Keynesian view, declaring the virtues of "aggressive monetary policy" to fight recessions, and making the case for discretionary fiscal policy too. (Naturally, the only form of discretionary fiscal policy considered was tax cuts, but the logic was straight Keynesian, and could have been used to justify public works programs equally well.). Oh, and the report - presumably written by Greg Mankiw - used the "s-word", calling for "short-term stimulus". [Krugman]
Greg Mankiw, economic adviser to Bush Jr and presidential candidate Romney, is one of the leading 'new Keynesian' economist, a form of Keynesianism that tried to reconcile Keynesian macroeconomics with traditional microeconomics (the latter based on maximizing individuals and firms, but let us not bore you with the technical details here).
So it can be argued that the outright hostility to Keynesianism, and especially the volume of it, is a relatively recent phenomenon, which in itself warrants explanation. But let us first give a short description of what Keynesianism is, and why it is still a particularly useful perspective to use in assessing our current economic condition.
After a long gestation period, Keynes wrote his major work (The General Theory of Employment, Interest and Money) in the Great Depression of the 1930s. As so often with great thinkers, it's a rich, but also very complex work that has sprouted many different interpretations. Although Keynes himself made something of a caricature of the 'classical' economics from which he tried to distinguish himself, he was right that the latter had no convincing theory of why the Great Depression happened, and their policy prescriptions were often counterproductive.
With the complexity of Keynes' work, and his untimely death in 1946, various flavors and schools of Keynesianism emerged. The (technical) differences do not really have to concern us here. What's more important is what unifies them, we'll dumb it down to its essential core:
- Economies are prone to significant fluctuations and can be stuck into underemployment equilibrium for prolonged periods of time, that is, they can become demand constrained with many resources lying idle.
- Therefore, some form of counter-cyclical policy is necessary.
Unlike what some seem to think, Keynes wasn't anything remotely resembling a Marxist, in fact, he believed he was saving capitalism from its own defects and thereby preventing more radical solutions (needless to say, in the 1930s these radical solutions were rather more manifest than they are today).
Arguably, this is the central concept of Keynesianism (it is for us) and many would agree that it's the true innovation of Keynes. While Keynes did make something of a caricature of the classical economics that went on before him, it's no exaggeration to say that these more or less all struggled to explain something so momentous as the Great Depression of the 1930s.
In classical (and it's modern version, new classical) economics, markets more or less clear instantly and production creates its own demand ("Say's Law"). The flip side of that is that it becomes rather difficult to explain the ups and downs of the business cycle, let alone something bigger like a depression.
While there are many different theories and strains within Keynesianism to explain exactly how an economy can end up being demand constrained for a prolonged period of time, the details do not necessarily have to worry us here as these are blatantly obvious in the Great Recession of 2008.
When house prices collapsed, the debt that financed the housing boom remained, wiping some $9 trillion of households' balance sheets and many lost their houses altogether. In order to repair their balance sheets, households increased savings and reduced borrowing and spending.
This reduction in spending in order to repair balance sheets compounded the shocks from construction and finance and created some sort of a chain reaction, which led to a really sharp downturn in the economy (sinking over 9% at one stage). Since my spending is your income, and vice versa, these downturns can feed easily on themselves.
Policy saving the day
Which is why the Cavalry stepped in, first the Fed saved us from a melt-down in the financial system, then the Federal Government bailed out the auto industry and came with (a rather insufficient) $800B stimulus program. There is a remarkable amount of people that argue that these actions didn't help, and we have to admit we're a little astounded by that.
If you read back how the financial system froze up after Lehman Brothers was allowed to go bankrupt, then it becomes really difficult to imagine that there are people (many Austrian economists and pundits) out there that argue that we should have let them all go bankrupt. The modern financial system is extraordinarily complex (tons of lawyers and forensic accountants are still sifting through the exact obligations of Lehman) and interconnected.
And, at least at the time, highly leveraged, very highly leveraged (that should never have been allowed, but that's another story). Basically, the authorities wanted to set an example with Lehman. Well, we know the results. If one bankruptcy could have brought the whole financial system to its knees, imagine what would have happened if they had let them all go bankrupt.
More controversial was the Federal stimulus bill. Academic research, on the whole, argue it was successful, but there were problems:
- An underestimation problem, the downturn was much worse than the government realized when the stimulus was drawn up
- The result of that was a communication problem, the government argued it would keep unemployment from rising above 8%, it didn't, and much of the public therefore regarded the stimulus as a failure
- It wasn't the most effective plan. Half of it were tax cuts, which, under the circumstances of rapidly worsening balance sheets and economic conditions, were likely to be saved rather than spent.
So popular folklore has it that it was a waste and many critics of Keynesianism argue that we should not have intervened. But the evidence of the success of these policy interventions is pretty conclusive. The economy was in free fall, until something stopped that and reversed the situation:
Yes, since then, the recovery has been sluggish, but there is a wealth of evidence pointing to demand constraints as the main problem, evidence to which we're turning now.
Above you see that capacity utilization of U.S. industry, after taking a big hit in the 2008 crisis, is still below a more normal 80% level. If present capacity is under-utilized, companies have less incentives to build new capacity, and that in itself restrains demand and leads to less productive capacity:
In the figure above you see that the 2008 crisis opened up a big gap between actual GDP and potential GDP (the latter is what GDP would be if all productive factors would be employed), this is the so called output gap. It clearly shows that the economy is demand, not supply constraint, although the output gap is declining.
The figure also shows the trend output growth, and there is a growing gap between trend output growth and potential output. This is simply what a reflection of what we already just mentioned, an output gap decreases the incentives to build new capacity. As a result, less production capacity is added, hence the growing divergence from its pre-crisis growth path.
The longer a material output gap remains, the more future capacity is reduced to what it could have been if there would not have been an output gap. What's more, less investment in new capacity also means less introduction of the newest technologies, not only is the productive capacity smaller, it's also older and less productive:
Capital expenditure relative to sales is at a 22-year low and some strategists reckon the typical age of fixed assets and equipment has been stretched to as much as 14 years from pre-crisis norms of about 9 years. [Reuters]
Labor is also under-utilized:
The total "jobs gap"-the number of jobs needed to return the U.S. economy to pre-recession health-is 7.9 million jobs (3.6 million for women and 4.3 million for men). [Heidi Shierholtz]
Most of this is a reflection of the sluggish demand and investment and demographics almost certainly explains a considerable extent, but just how much, well, opinions differ:
Economists disagree, however, on exactly how much demographics are responsible for the current fall in the participation rate. The Chicago Fed estimated in 2012 that retirements accounted for one-fourth of the drop in labor force participation since the recession began. Other analysts, including Barclays, have suggested that aging Boomers could account for more than half the drop. [Washington Post]
What's not in dispute is that labor is significantly under-utilized and that a revival in demand could easily put lots of people back to work. The sooner that happens, the better, as long-term unemployed tend to lose skills, motivation, and are discriminated against leading to a destruction in human capital (apart from the human tragedy involved).
Saving - Investment balance
Another piece of evidence comes from sectoral balances. We already described what the 2008 crisis did to household balance sheets, and how households responded by cutting back borrowing and spending, and increased in savings in order to repair the damaged balance sheets.
With a reduction in demand, companies also cut back on investment, despite record low interest rates. Combining these reactions in one graph, you get the S-I curve, or the private sector financial balance. You see in the graph below that this sharply moved into surplus, with S increasing and I declining.
But you see something else, the uncanny way unemployment is correlated with this graph. That is, a retrenchment in private sector spending clearly goes hand in hand with a rise in unemployment. Now, correlation doesn't predicate causation. But we have a simple and adequate mechanism for that, which we already described above. A cutback in spending produces a cutback in production and investment, and a rise in unemployment (leading to further cutbacks in spending), the simple Keynesian multiplier.
We think this is far more likely than any of the alternatives. In a neat classical model, a cutback in spending leads to instant wage and price adjustment, keeping all resources fully employed. This is very neat in a model, and we understand its lure, but reality simply doesn't work like this.
Real business cycle models (also in the classical, market clearing, tradition) have no bearing on the 2008 crisis, which obviously wasn't caused by any random technology shock. The Austrians argue that we should just let it all play out without policy intervention, supposedly to get rid of all the artificial capacity that has been built on credit that was too cheap, but this is oblivious for Fisherian debt-deflationary spirals.
Speaking of different economic models, this is another, and perhaps the strongest, level of evidence in favor of the Keynesian nature of the crisis (that is, output being mostly demand constrained). As Krugman has shown time and again, the simple Keynesian IS-LM model he uses has been able to explain a host of data, which is vexing alternative models:
- No take-off in inflation despite massive amounts of QE (or 'money printing' as the critics have it)
- No dollar crash, despite massive amounts of QE and record low interest rates ('currency debasement' as the critics have it)
- No spiraling out of control of bond yields despite record levels of public deficits and debts
Krugman already wiped the floor with these predictions in the immediate aftermath of the crisis, and he was simply basing his stance on his simple Keynesian IS-LM model, arguing with the likes of Niall Ferguson who worried about rising bond yields in 2009(!) and Allan Meltzer on inflation in 2009. Or that alarming open letter to the Fed in 2011.
One might also want to consult his "Dark Age of Macroeconomics" in which he takes on Eugene Fama and John Cochrane for making débutant errors in arguing bond financed public spending will crowd out an equal amount of private spending.
In short, that simple Keynesian model has been able to explain stuff that is vexing opponents. And there is more, here is Richard Koo, from Nomura:
If central banks bring interest rates down to almost zero and nothing happens then it is not an ordinary world.
Indeed it isn't. This is the strange world of a liquidity trap, an extreme case of a demand constrained economy. We have already explained the (often counter-intuitive) mechanics, but this is a world Keynesians are familiar with, while it vexes other schools of thought.
The only one who, to our knowledge, has publicly admitted there might be something wrong with their economic world-view is Art Laffer, a supply-sider, but we guess it's difficult when so much of one's work, reputation, and perhaps even identity is tied up to one perspective to suddenly confess that it can't explain today's world.
Public sector deficits and debts
Apart from those bond yields not rising in the face of massive deficits and debts, Keynesians never really overly worried about the deficit and debts, at least not in the immediate future. When the private sector is deleveraging (spending less to repair balance sheets), the worst thing to do is compound that by public sector cutbacks. We actually know that from history, like the U.S. in 1937, Japan in 1997, or what is going on in the eurozone periphery today.
If you wonder why the recovery has been so tepid, here is an indication:
A key driver of this broad-based weakness is austerity in the public sector. Since the recovery began in June 2009, the public sector has lost 728,000 jobs. However, to keep up with population growth over this period, public-sector employment should have increased by around 750,000. That means the total gap in public-sector employment today is around 1.5 million jobs. Nearly 30 percent of that gap has occurred in local government education, which is mostly public K-12 employment. [Shierholtz]
There has also been a re-assessment of the size of the multiplier on public sector expenditures under these circumstances (in which austerity cannot be offset by lower interest rates as these are already zero). The IMF has calculated that the multiplier could be up to three times as large as previously thought, so austerity is rarely a good idea under these circumstances and can easily become counterproductive.
And yes, the budget deficit was scary at one time (which was used to great effect), but it was in large part cyclical, that is, the result of the private sector deleveraging itself and really not that scary.
And then there is simple survey data, showing how the crisis affected borrowing, spending and risk aversion. An Associated Press survey showed what consumers did with their money in the five years before the Great Recession began in December 2007 and in the five years that followed, through the end of 2012. The focus was on the world's 10 biggest economies:
- A retreat from stocks
- Shunning debt
- Hoarding cash
- Spending slump
We'll give you a few quotes from the report:
The growth of cash is remarkable because millions more were unemployed, wages grew slowly and people diverted billions to pay down their debts... A flight to safety on such a global scale is unprecedented since the end of World War II.... People chose to shed debt even as lenders slashed rates on loans to record lows. In normal times, that would have triggered an avalanche of borrowing.
You'll get the picture.
While Nixon could claim in 1971 that we were all Keynesians, during the stagflation of the 1970s, Keynesianism lost some of its luster, as it was supposedly unable to explain the stagflation crisis. This is simply nonsense. The 1970's stagflation crisis was a supply crisis, not a demand crisis.
There was a combined oil and commodities shock, and (more importantly) a secular reduction in the growth rate of labor productivity. The results can be simply shown as a leftward shift of the aggregate supply curve, leading to lower output and higher prices, voilà; stagflation.
There was criticism from Friedman and Phelps, arguing that there is no long-run trade-off between unemployment and inflation (that is, the long-run Phillips curve is vertical). Indeed. But this doesn't invalidate Keynesianism as demand measures only make sense when the economy operates way below capacity and unemployment is above the natural rate, with which the Friedman/Phelps result isn't incompatible. They argued using demand to push unemployment below the natural rate will only work in the short run, in the long run it just leads to higher prices.
Another theoretical argument directed at Keynesianism has been that it lacks 'microfoundations,' that is, it's not founded on the rigorously rationally optimizing individuals we know from microeconomics. While that has been remedied to a large extent from the 1980s onwards by the so called new Keynesian school, we were never very impressed with this criticism in the first place.
While some of the assumptions of traditional Keynesianism (like nominal rigidities) seem ad-hoc, empiric is firmly on their side. And we simply know that individuals aren't nearly as rationally optimizing, nor that markets are perfectly competitive, transparent, always clearing, never trading when not in equilibrium, all firms having equal access to the same book of technologies, the same upward sloping supply curves etc., etc., in short, the host of unrealistic assumptions made by standard microeconomics.
So the battle for microfoundations was one of empirical foundations versus theoretical purity. That was always a bit of a no-contest for us. We're firmly on the side of empirical relevance.
Many reactions to some of our articles seem to think that all we ever want to do is to spend, spend, spend. No. It's horses for courses. When the private sector is deleveraging and/or the economy is demand constrained, spending is the obvious thing to do, especially if monetary policy lost much of its traction (liquidity trap).
However, if an economy is producing more or less at full capacity, then it obviously isn't demand constrained but supply constrained. More public spending under these circumstances doesn't make sense and is highly likely to be counterproductive (as Friedman and Phelps showed).
If in doubt, ask the Argentines, as that is exactly what they have been doing since 2007, with all the predictable outcomes (inflation, interest rate, currency collapse) that Keynesian critics predicted for the U.S. And if and when output in the U.S. approaches full capacity, that criticism will hold, and we will change our diagnosis, opinion, and remedy, but not before.
Some even argue that stimulus might be necessary under certain conditions, but it's never being withdrawn. There is simply no evidence for that. It's certainly not what happened with the 2009 stimulus bill.
So, we have:
- Keynesian interventions that worked in the 2008/9 crisis (and in the U.S., Japan in the 1930s, etc.)
- A host of evidence that the economy is demand constrained, from output gaps, unemployment, sectoral balances and survey data
- Simple Keynesian economic models being able to explain a host of phenomena other approaches struggle with, mainly because they fail to recognize the economy is demand constrained.
We fail to see why Keynesianism engenders so much outright hostility. Its diagnosis (a demand constrained economy) was right, supported by a host of evidence. Its model predicted beautifully where others failed. Its policy recommendations, where implemented, worked. The empirical foundation of Keynesianism, that economies can produce below capacity for prolonged periods of time, and self stabilizing forces can be weak or even overpowered by self-reinforcing mechanisms is sound.
No, it's not a panacea. Economies are not always demand constrained, nor is that necessarily their only, or even main problem. Hence us arguing in favor of big, immediate tax cuts in France, or why we would argue Argentina should make budget cuts and stop printing so much money.