Friday, May 27, 2011

The structural differences between the USA, Europe and Asia

There is much written these days about the big economies of America, 'Europe' and China, and the differences between them.

What is curious about this is that I haven't found anybody who has taken the time to actually visualise the differences.

This can be done a couple of ways and I will hopefully do both over coming weeks. The first would be to look at the differences in production co-efficients ( the ratios of inputs to make a $ of output). Alternatively, you can visualise the intermediate goods matrix. To do the latter I suggest the folowing.

1. I think you need to first develop a panel GDP set for both time series analysis and for international analyses so that the comparisons are actually on a dollar for dollar basis. It is most common to use national GDP as the denominator for comparisons but I think we should be doing better that than.
I made the argument here for panel GDP analysis . In a future blog I'll some international R&D analysis that I have developed to compare GDP with Panel GDP.

2. I have created an OECD GDP Panel dataset for fifteen countries starting in 1970.

3. So that done you can visualise the economies using the landscape chart format that I have explained previously in this blog.
The data comes from the OECD input-output database for 2005 and the categories on both axis are simply those from the I-O inter-industry matrix (i.e. not the value  added of final demand sectors).

Click on the images below to enlargen.

1. The USA 2005.

What we see from this image is that the US economy's interdependencies are heavy weighted to the services sector.

The super tall peak is the interaction of the supplier industry 'Finance and insurance' and the demand industry of 'Finance and insurance' . In I-O economics it is most common that the intra-industry interactions are the largest.

2. Europe 2005.

We see that Europe (all the EU countries in the database) is also heavily weighted to the services sector. Now this image is not as accurate as it is constructed from the sum of domestic tables. It thus leaves out cross border but intra-european trade - which will favour manufacturing interactions. In future I'll post historical analysis of Europe that is more accurate as I have international input-output matrices for 1970, 1990 and 2000. For anything later than that will wait on the work of the European funded WIOD group.

3. Asia.

For the Asian countries (China, Japan, Korea, Indonesia and Taiwan) the same problem as applies to the Europe chart apply. It nevertheless emphasises the huge structural differences between the economies of America, Europe and Asia. Asia is heavily concentrated on manufacturing.

So, despite the inaccuracies, these charts provide a very interesting insight, though rough approximations of the relational structure of these three 'economies'.

The big discussion point out of these charts is that they reinforce the perception that services are based in the west and manufacturing in Asia.