A monumental data gathering effort

is being led by Peter Nardulli at the University of Illinois at Urbana-Champaign. As I mentioned in an earlier post, it involves the digitization, classification, and coding of 30 million plus news reports from all over the world going back to 1946.  Called Event Analysis, the project will greatly enrich the quantitative information at social scientists' fingertips. You want to know how frequently labor protests accompany IMF missions to developing countries? Whether governments announce more new programs before or after elections? How quickly economic or political ideas and fashions travel across countries? A few clicks of your mouse, and you will have your answer to these and a million other questions about recent economic and political history. 

Two downsides: it will put a lot of graduate students out of a job (no need to pay them to collect and code information once there is a master data set), and it will probably lead to too much data mining.

Comments

[Nilgun]

Brilliant... Thanks for sharing this information with your readers too. I always find something precious whenever I check Dani's weblog.

[Murat K]

It might also mean better use of grad students' skills, at least for some of them.

Am I too optimistic for a grad student?

[Francis Bacon]

This is incredible. Maybe if aid organisations (both state and NGO) released all the information they also owned, something similar would be possible in the development sector?

For more, see:
http://thatsthewaythemoneygoes.blogspot.com/

[jeremy hunsinger]

hmm, this sounds something like the work that my colleagues are ohio, virginia tech and harvard were doing based on the World Handbook of Political and Social Indicators work.