Edward Glaeser has written a couple of posts over the past week looking at Argentina on the Economix blog for the NYT. It’s an interesting topic. Argentina was a wealthy country at the start of the 20th century, but by almost every conceivable ecnomic measure available, it is much worse off now than it was then. Here is an example from my own research:
This shows how well connected a few countries have been in the international trade network from 1938 to 2003**. The diagram shows several of the common patterns in the data. The US is the line in the middle – it has always been the 1st or 2nd most connected country in the network. The lines for Singapore and China are moving towards the middle, indicating that they have become increasingly more connected over time – China is now one of the unusually-highly connected hubs within the network. Ireland is one of the countries that has drifted up and down in the rankings, and Argentina is one of the countries that has dropped dramatically – from 9th most-connected in 1938 to 98th most-connected in 2003. This drop in trade connections correlates quite strongly with relative drops in GDP per capita too.
In his posts, Glaeser has been trying to tease out the reasons for this decline. In the latest one, he includes this graph showing the correlation between school enrolments in 1900 and GDP per capita in 2000:
This data is astonishing! It shows an incredibly strong correlation between school enrolments in 1900 and wealth 100 years later. Glaeser goes on to discuss whether this is a random correlation or suggesive of causality (and he comes down on the side of causality). This link between education and wealth has several important implications for those of us interested in innovation.
The first is that it offers more evidence that the roots of innovation are deep. The work of David Landes and Nathan Rosenberg among many others shows that differences in innovation performance at the national level explain differences in GDP measures. The decisions that we make now about education and other infrastructural support for innovation will have an impact for many years to come. Just as a firm can’t simply say ‘we’ll start being more innovative right now’, neither can a country or a region.
The second point is that this shows part of why there might be a link between innovation and wealth. Again, there is plenty of research that shows that educational levels correlate with innovation. I would contend that since innovation is primarily the act of making novel connections between things, a large part of this link comes from the increased ability to make interesting mental connections that is part of a successful educational process.
It also makes an alarming point about politics. If the impact of decisions about things like education continue for a century, then policy-making based on the 24 hour news cycle seems especially dangerous, doesn’t it?
I think that the case of Argentina tells us a lot about innovation policy. In particular, the importance of thinking through the implications of what we’re doing right now for what we want to have happening in the future. Even though very few of us have 100 year planning windows, this is still an important point.
**Specifically, this data ranks all of the countries included in the IMF trade data over that time based on their in-degree within the network. In-degree is a measure of the number of countries that export a significant amount of their GDP to a particular country.