Here is year-over-year growth in capital formation in about 45 countries. Capital formation is investment and in most cases is dominated by construction. Some of the data is nominal, some of it is real. I actually think the discrepancy helps my point.
The data starts in 1960 and comes up to the present day. It really looks like something “different” and global happened in 2008. There is a lot of noise and some individual countries moving way off track. Also new countries are added to the set as time goes on.
However, there is a near universal nick in the series in 2009. The only country not fully nicked was Indonesia, the dark green line spanning the gulf. The slowdown in capital formation was only slight there. We could even say that the importance of oil in Indonesia “explains” its outlier status, but I don’t even want to go there.
Even with Indonesia its really looks like something unprecedented happened in 2009 and it happened everywhere and it had to do with capital formation.
Its not clear how we can explain this with any factors that have to do with local monetary policy or inflection points in technology. Maybe we could say that the interruption in Patterns of Trade and Specialization was so profound that it impacted everyone.
That seems a stretch, its not clear why such a decentralized process could have such a universal effect. It would almost be like a global PSST virus that shutdown lots of connections at once.
7 comments
Comments feed for this article
Monday ~ June 6th, 2011 at 11:40 am
Assorted links — Marginal Revolution
[...] 4. What has happened to global capital formation? [...]
Monday ~ June 6th, 2011 at 6:04 pm
Leigh Caldwell
Could this graph be biased by measurement error?
If (as you point out) the figures are dominated by construction, could this be because it is the only form of capital creation that is easily measurable by national statisticians? If so, then a fall in construction would be easy to mistake for a fall in capital formation of all kinds.
A fall in construction could quite easily be caused by a coordinated mistake – investors globally in 2004-07 coming to believe that general growth would continue smoothly, and therefore demand for property would rise correspondingly. The worldwide interruption in growth quite rationally results in a smaller optimal total housing stock.
The nature of construction in particular, and to a lesser extent capital in general, is that a small change in expected growth leads to a large change in the optimal rate of capital investment. That’s assuming growth is on a unit root path instead of a reversion to trend.
Monday ~ June 6th, 2011 at 8:42 pm
Lord
Yes, but a credit shock could be monetary (tight money) or it could be real (failed profit expectations) or both. Could the expectations people built up during the boom have ever been fulfilled? I doubt it. They only could have been met with earlier disappointment but at less cost.
Friday ~ June 10th, 2011 at 4:37 pm
Greg Ransom
Karl, you’ve no doubt heard the term “Austrian business cycle theory” but have you actually studied the work of Hayek or Garrison or Horwitz or William White?
I’d particularly recommend Hayek’s _Monetary Nationalism and International Stability_.
Hayek’s business cycle theory is international in scope, even if those ignorant of his work have no way to imagine it.
Friday ~ June 10th, 2011 at 4:38 pm
Greg Ransom
This is just the phenomena Hayek’s work provides a causal mechanism for explaining:
“its not clear why such a decentralized process could have such a universal effect. It would almost be like a global PSST virus that shutdown lots of connections at once.”
Friday ~ June 10th, 2011 at 5:26 pm
ezra abrams
what a crappy blog post
Did the author actually look at his own data ?
without even bothering to mention the poor layout and design of the graph, it is not at all clear the 2008 is that odd
I suppose, if the author actually bothered to sift thru the data, and present it in some fashion that was interpretable, he might get a critical reader to agree with him.
But as the blog post stands, I don’t think any one with a critical mind can look at the data and find anything, other then that in around 1993, the country with the salmon line spiked (which is a clear example of the poor quality of the graph: you can’t tell what country it is, and even if the authore had bothered to write out the names of hte countries, with that many colors, you cna’t tell what color in the labels goes with what color in the graph…almost a text book case of someone who needs to re read tufte)
Saturday ~ June 11th, 2011 at 1:45 am
J Thomas
Ezra, I agree the graph is not great. But it should be obvious that the salmon line is Turkey.
And it should be obvious that something special happened around that time. In 1983 some were a bit down and some a bit up. In 1992, some were a bit down and some a bit up. In 1998, some down and some up. In 2008 a few were up sharply and some down, and by 2009 everything was down. I don’t know why you’d deny that the graph shows this, imperfect as it is.
Maybe there’s some simple obvious explanation. Obviously, national economies are interconnected. I don’t get much out of somebody saying “My economic theory says this is possible so that makes my theory special”.