It is approaching ten years since the collapse of Lehman Brothers. To most people this is when the great financial crash started, since it is from this point that the most serious repercussions started to take place. Personally I date it from the collapse of interbank markets more than a year before, but never mind. What was so shocking about the crash was that so few trained economists foresaw it. Or, lest anybody mistake that for mere criticism of forecasting skills, how so many economists aided and abetted the forces that destabilised the financial system, or at any rate, did nothing to head the crisis off. Ten years on and I am starting to wonder whether we are unlearning the lessons of the crash, and relearning the complacency that let it happen.
What shook me was an article in the Financial Times by long-standing correspondent Martin Sandhu – the devasting cost of central banks’ caution. His suggestion is that central banks were too cautious in the aftermath of the crash, and as a result the recovery has been stunted. What struck me about this article is how heavily it is based on the economic conventional wisdom that helped cause the crash in the first place. It is as if economists like Mr Sandhu had learnt nothing at all. In particular it is based on three persistent fallacies that should by now be well and truly exposed.
Fallacy 1: the pre-crash trend growth rate was real and sustainable
Economics is a numerical discipline, and economists are more than usually susceptible to the delusion that numerical measures are more real than the complex world they describe. One case in point is something called the “trend growth rate”. This is an average rate of growth taken over a long period of time. By its nature it does not vary much from year to year, but before 2008 it had been remarkably consistent in developed economies, at about 2% per annum. After the crash it fell, and a new, much lower trend has established itself. Mr Sandhu’s contention is that the pre-crash trend was ordained by the laws of physics, and that the decline must represent a policy failure – which he lays at the door of central banks. The gap between where the economy is now, and where it would have been had the pre-crash trend persisted, is now very big – and it is the “devastating cost” of the article’s title. This is a persistent idea, which I have heard from Mr Sandhu’s FT colleague Martin Wolf too, though I suspect that he is starting to wake up to the truth. You also hear it from some left wing commentators who lay the blame at the door of “austerity” rather than central banks.
The persistence of the trend before 2008 is all the evidence that many macroeconomists need that it must represent some natural law: the steady advance of technological change on productivity. They did not think it was their job to ask more penetrating questions. But the cracks were showing well before 2008, and had economists spotted this, they would have seen that instead of the world continuing on its steady course, the risk of an economic crash was building up. The crash blew away froth that had been accumulating in the years before.
There are several aspects to this, and I will try to be brief. First is that the sort of productivity growth that economists fondly believed is happening everywhere, was always confined to fairly narrow sectors of the economy, and those sectors are themselves narrowing. In particular it applied to agriculture and manufacturing: industries where you produce stuff and ship it. And yet the importance of these industries to developed economies is steadily diminishing, in a phenomenon called the Baumol Effect that is taught in Economics undergraduate courses, and which should have been no surprise. The true drivers of a modern economy are in services such as healthcare, where technological advance does not tend to advance productivity in the sort of way that is captured by economic statistics.
Second, in the ten years up to 2008 much of the rate of growth depended on globalisation, and in particular the use of cheap labour in the developing world, notably China. This was strikingly evidenced in the fact that the prices of many manufactured goods actually fell in the period. This was not a case of advancing productivity: it was the exploitation of another basic economic idea whose implications few modern economists actually seem to understand: that of comparative advantage. The problem is that where this arises because the developing world is catching up with the developed world these gains are temporary and actually go into reverse eventually. That was what was starting to happen before 2008 as Chinese wage costs relative to American and European ones started to close. It was not sustainable.
And third, much of the growth, especially in Britain, depended on industries whose economic benefits were actually doubtful, and where productivity measures are similarly flawed. Two sectors stand out. One is finance; not only are the actual benefits to human well-being from its output hard to understand, but much of the reported income turned out to be downright false. That is the main reason why the crash had such a drastic effect on GDP figures in 2009, as bank bailouts were needed to replace non-existent assets created by this fictitious income. That should be reasonably obvious, though I’m surprised by how few commentators pick this up. More subtle is the problem with a second sector reporting productivity growth: business services. This is things like lawyers, accountants, consultants and architects – particularly important in Britain. The problem here is that these do not provide services to end users but to other businesses. Which leads to the question, how can you say that these industries were becoming more productive when there is so little evidence of productivity gains in the industries they were advising?
Fallacy 2: consumer price inflation is the main evidence of unsustainable growth
In the 1970s developed economies overheated, and this led to an inflation price spiral, as consumer prices and pay rates chased each other. This wasn’t supposed to happen under the Keynesian economic regime that most of the world used at the time, because unemployment was increasing too. This was a huge shock to the economics profession (a much greater shock apparently than the 2008 crash), which led to a brand new policy framework. This led to an obsession with consumer price inflation. If you keep this at some Goldilocks level, of between 1% and 3% per annum, then nothing too much could be going wrong, people thought. This was made the sole performance target for many central banks, as it meant that economies were neither overheating nor underperforming. Other things, like the level of bank lending, asset prices and balance of payments imbalances didn’t really matter. This was central to the neo-Keynesian consensus.
If this was ever valid, it was rendered obsolete by the advance of globalisation in the 1990s and 2000s. Wages became detached from consumer prices. Capital flowed freely between the world’s economies. Especially in Europe the movement of labour from one country to another became freer. There were many ways an economy could overheat without consumer price inflation becoming affected. Asset prices (real estate, shares or even bonds) could inflate into bubbles; dangerous levels of unreal assets could build up within the financial system; a destabilising level of migrants could cross borders. All of these things happened in Britain in the run up to 2008, and yet because inflation remained in the Goldilocks position people thought everything was fine, and that all these other things would just sort themselves out. Many still believe this; it is one of the few things that unites all factions of the Labour party, which was in power at the time. Likewise Mr Sandhu persists in suggesting that consumer price inflation was all that central bankers needed to worry about after the crash – and because it has shown no sign of taking off, there has been plenty of scope to pump things up.
Fallacy 3: monetary policy is an effective way of regulating demand
This was another part of the neo-Keynesian consensus. Prior to this the conventional wisdom that fiscal policy (taxes and public spending) was the best way to help a flagging economy or cool an overheating one. But after the failures of the 1970s it was felt that this got too tangled up in politics to be effective, and would lead to an excessively big state. Instead the job could be given to central banks, through changes to interest rates and other things like Quantitative Easing. These things were referred to as monetary policy, conjuring up the quaint image (not discouraged by economists) that it was about the printing of banknotes.
And yet this idea has not stood up to the test of time. Quite apart from the fallacy of using inflation as the main performance indicator, monetary policy affects many other things than consumer demand, and in fact has led to a build up of dangerous levels of private debt. The political pressures to keep interest rates low, especially if inflation is low, quickly become unbearable. And not just political pressures. Increasing interest rates can destabilise the banking system. Such increases may have provoked the pre-crisis in 2007 that in turn led to the main crisis. Monetary policy has no brakes.
Here the left wing critics of economic policy, who think that fiscal policy should do more of the work of regulating demand, are on stronger ground. But would politicians have the guts to tighten fiscal policy when needed? The vehemence of left wing criticism of austerity, and their denial that the pre 2008 economy was overheating, suggest that fiscal policy too lacks effective brakes.
Are we heading for a new crash?
There are some worrying signs, with the build up of private debt, and with reckless fiscal policy in the US. But we are not seeing the same reckless advance of “financial innovation” that proved so devastating ten years ago. Not in the developed world anyway: China looks a bit different, though the awareness of the central authorities there is a strong contrast to the denial of many western leaders before 2007.
But some trouble is surely ahead. Perhaps then economists will start to break out of their bunker and start viewing the world as it really is, instead being a construction of outmoded statistical aggregates.