The last post here received an unexpected amount of attention, so we thought we’d do some further work on what is, and is not, linked to the government cost of borrowing within the Eurozone. We found some more surprises…
About the Authors of this post:
We discovered that a model of GDP per capita and use of broadband (our alternate measure for a country's internet use) was extremely successful in determining the government cost of borrowing.
We found that it is not enough just to be a rich and productive country (as measured by GDP per capita) to have low borrowing costs. This must be matched with high quality (fast), widely adopted, and low cost broadband to have the lowest cost of borrowing possible in the Eurozone.
That is to say, the more productive the economy and the greater the country’s broadband capability, the less risky the Eurozone country's economy.
This matrix shows countries in the groupings determined by their relative scores in broadband infrastructure and GDP per capita. The numbers in the coloured boxes are the average cost of borrowing for the group over a one year period.
Readers of the previous blog will note that we have dropped the BCG e-intensity index. The e-intensity index is a closed box (to us) as we don’t know precisely what goes into it. It is also correlated strongly with gdp per capita – meaning importantly that it is a good measure of productivity and modernity, but does not necessarily separate the two.
We have replaced the BCG's measure of e-intensity with one calculated by the ITIF (Information Technology & Innovation Foundation) that transparently (and reproducibly) combines adoption, cost and speed of broadband within countries. Doing so enables us to separate out the effects of GDP per capita and those of Broadband.
It seems commonly accepted that factors such as the current account deficit and public debt as a % gdp explain bond yields. We found the relationships were nothing like as strong as those of GDP per capita and Broadband, and their inclusion added no additional significance to the analysis.
We have tried several other factors as well as those above and failed to find any that provide as strong a link to Eurozone bond yields than in our proposed model. Not being able to find significant explanatory power in the commonly accepted factors leaves us with the view that there is something substantial in our simple analysis.
Although correlation does not imply causation, it seems to us that broadband and its use plays a far bigger role in modern economies than its share of GDP might suggest.
What next? To go much further you need more data. Demonstrating a causal relationship would be extremely interesting. Part of such an analysis would be to go back 8-10 years and track the evolution of Broadband and measures of country stability, inflation and productivity – we suspect that this would also reveal insights as to how the internet-ification of countries changes their capacity to deal with shocks.
In summary, we have modified (explored) our original analysis and the message is startlingly similar: those countries with good broadband and higher GDP per capita have low cost of government borrowing for any given level of public debt.
About the Authors of this post:
David Cleevely is Founding Director of the Centre for Science and Policy at the University of Cambridge.
Matthew Cleevely is an entrepreneur pursuing a PhD in entrepreneurship, innovation and growth policy at Imperial College Business School, London