Graphs!

Not so long ago I responded to a post on Mark Kleiman's weblog concerning the "competitiveness leagues" that (usually) rightwing organisations like to prepare. I produced some methodology questions and some questions about content, here. I also proposed to check the performance of one of these against results, and in the end I did.

The one I looked at was the "world competitiveness ranking" published by the Swiss Institute of Management Development, who can be found here. Figures were available for the years 2000-2004, for some 60 countries and regions. Not all of these, however, ran through the whole period or were comparable with the OECD growth stats, in the OECD Economic Outlook No.76. I decided to plot the GDP growth rate on one Y-axis and the ranking on the other, with time on the X-axis. (GDP figures are in real terms, annual percentage change.)

So, how did it turn out? Well, the first point must be that it worked rather better than I expected. The chart for France, for example:graph

Well, that seems clear enough. The chart for the UK isn't bad either, but it does have some noticeable outliers - 2001's rise up the ranking is reflected by a corresponding slowdown, and 2002's fall in the ranking mirrors a noticeable acceleration in growth, although overall it seems quite good. Australia's chart is well correlated, too.

It's not that simple, though: try this chart for the Czech Republic. If there's a correlation here it's weak. The charts for Poland, Turkey, Denmark, and Canada, though, are nowhere near as good. In fact they show quite a marked tendency to go in the opposite direction to the growth figures - the curves for Canada, in particular, look like mirror images of each other. Turkey's ranking fell gradually over the period of the study, but after a sharp recession in 2001 (no doubt due to the currency and political crisis that year) its growth rate soars away. Japan's is rather hard to read - although the correlation seems very strong, this is only part of the story. The change in Japan's ranking seems to track the business cycle, but in absolute terms the ranking is very high for an economy producing not very much growth.

Conclusions from this fairly unscientific review? Better than I thought, but falls down pretty badly on quite a few cases. I think I may also compare some of these with the aggregate figures (OECD and world growth rates), as I suspect some of the correlation is with the business cycle.

And will some of you comment on this? I can't work out why, but people seem to prefer the sarcastic abuse and international gun-running to the economics posts.