Almost ever since I had learned about the Austrian Business Cycle Theory (ABCT) and the details of the US financial crisis and its background, I have felt that there is something wrong with attempts of many Austrians to prove that the less-than-prime mortgage bubble in the US that is widely regarded to have caused the subsequent severe stock markets crash and the subsequent recession is an instance of the pattern of inter-temporal distortion described by ABCT.
My main doubt concerned the fact that single-family houses with regard to which the mortgage boom took place do not seem to conform to the notion of an excessively roundabout investment projects the presence of a cluster of which on the eve of the crisis is the central prediction of ABCT. I am not an expert on how long it might take to build an average US single-family house but I doubt that it may take more than a year. Meanwhile, according to the widely popular version of events the housing boom lasted roughly from 2001 to early 2007 (see the graph below).
This is what inspired me into looking whether on the eve of the market crash there have been long-term projects abandoned en masse. Initially, I found a promising piece of information about abandonment in 2007 of 59 coal-fired power plants building projects. This did not turn out to be a breakthrough because when I looked into the details of the projects I found that the vast majority of them were abandoned at the pre-construction stage.
But while researching the fate of those projects I discovered a very interesting phenomenon where insiders of the coal power plant industry were explaining the difficulties they faced by in part a staggering increase around 2007 of the construction costs of their projects (see here and here). Then I found similar statements from the representatives of the US ethanol production industry (see here and here).
I then decided to look at the price dynamics of one of the most important input into almost any construction process – cement. If you look at the graph below (taken from this presentation) you may notice an interesting thing. The really substantial increase in the real price of Portland cement started only around mid-2004 and until then the price had actually been falling. At the same time the single-family housing boom was already in full swing at that moment. Something else may well have happened that could explain this discrepancy.
I will not make you wait for long. At least part of this something else seems to be commercial real estate which in the US also includes multi-family housing. Below you may see the price dynamics for commercial real estate (graph taken from here). You may see that they really take off around 2004, just a little bit before the take-off of the price for Portland cement.
But this is not all. If you follow the links to here, here and here you will quickly realize the extent to which starting at least from early 2007 the long-term commercial real estate construction projects have faced the difficulties in completion. We now have everything that is needed to plausibly collect the puzzle of how the Great Recession may, at least, in part of its severity, have been an instance of ABCT.
Of course the multiform incentives and direct and indirect subsidies for single-family housing construction have distorted the relevant market but that was, in my view, not conducive to a recession. After all, if we use a simple economic logic we should see that it is in itself difficult for a decline in consumption of a certain good (in this case single-family housing) to cause a recession because, although the producers of single-family houses may suffer, the would be buyers will retain the money to spend on something else.
It may be countered that the subsidies in themselves may cause malinvestment into the housing construction capacity and, later, effects similar to those described by ABCT. But this is implausible because the link between multiform incentives to get mortgages and their increase taking by customers was fairly obvious. The graph of cement prices that I have posted above contains potential indirect evidence for this idea. After all, the fact that the real price for cement did not substantially increase over the first years of the single-family housing boom but actually fell suggests that even the producers of cement managed to increase production sufficiently so as not to create significant shortages.
ABCT, on the other hand, explains the crisis by means of explaining how expansionary monetary policy of the central banks falsifies interest rates in the credit markets making some long-term investment projects appear more profitable than is warranted by objective market conditions (for a realistic explanation of how this process happens see, e. g. my recent paper) which in the middle of pursuing them turn out to be unprofitable as a result of unforeseen cost increases. The long-term commercial real estate is an excellent candidate for the cluster of such projects.
But what may this potential cluster have to do with the crash of the less-than-prime mortgage market that in its turn caused the crisis in the markets for mortgage-based securities and derivatives. The answer is that abandonment of long-term investment projects deprives people of income without freeing up the resources which have been wasted or deeply immersed into unsustainable investment projects. It is not unrealistic to imagine this leading to substantial increases in defaults on mortgages.
Of course, the picture I have briefly drawn in this post is a preliminary one. It may well turn out later that something in the logic does not hold water or that I have missed certain important factors and details. But it is exactly for quickly identifying such potential flaws that I am initially publishing this as a blog post.
Update: I found this short report in which it is claimed that overall construction spending in the US has fallen from $1,16 trillion to $803,6 billion, out of which the spending on commercial construction fell from $548,7 billion to just $231,7 billion. In other words, it seems that at least as of 2010, the residential construction spending has largely recovered (it was lower than the peak only by around $50 billion) while the commercial construction spending remained deeply depressed. It seems to confirm the hypothesis that large malinvestments were made in the commercial real estate industry prior to the recession.