Computer Class
You will be given long data series on military spending and unemployment for the UK and the US and use them to replicate published results for a simple two variable VAR.
Here is the exercise
Following Dunne and Smith (1990) "Military Spending and Unemployment in the OECD", Defence Economics, Vol. 1, pp 57-73, Estimate the simplest form of VAR to test for Granger causality between military spending and unemployment. Section 4 of the paper.
NB If you want to download the data in Microfit files press the shift key when you click on the names:
Use UR2 and GU, SM and GSM to replicate the results and use the Johansen procedure to develop the analysis. Is there a cointegrating relation between the two series? How might you develop this approach further?
Relevant pages of article are here, here here and here
Here is another exercise:
This following paper uses a cointegrating VAR analysis to investigate arms races between India and Pakistan and Greece and Turkey:
Dunne, J Paul, Nikolaidou, Eftychia, and Ron Smith (1999)"Arms Race Models and Econometric Applications". Chapter 11 in Paul Levine and Ron Smith (eds) (2003) "The Arms Trade, Security and Conflict", Routledge, pp 178-87. .
Replicate the results in the text using the following files. NB if you want to download the data in Microfit 4.0 files press the shift key when you click on the names:
India-Pakistan data
Greece Turkey data