When the facts change, I change my mind. What do you do, sir? JM Keynes

David McWilliams wrote a few months ago on the seemingly soaring number of GAA players transferring from Irish clubs to those abroad, linking the jump he saw in the data with the worsening labour market conditions for young men and women in Ireland. I thought David's analysis could use some work, and rebutted his claim in my own blog rant a few days later. I found that David had picked up a seasonal trend (low movement in Winter, lots of movement of players in Summer), so it didn't necessarily follow that the downturn in economic conditions was to blame.

Then I thought a little about it, and asked one of my Final Year Project (FYP) students, David Healy, to get the data from the GAA on player transfers for as far back as he could, and give the question a bit more of a look.

Here's David's 90 page FYP as a .pdf. It is 1.2mb, so make sure you are on a good connection.

David did a bang up job of getting, cleaning, coding, and analysing the data, and here's what he found. This is applied, timely, relevant, and high level data analysis from a UL undergrad. If you'd like to mess about with the data yourself, the data is here, zipped. The file is 10mb.

The idea was to take the ten years' worth of GAA transfer data David got from the nice people at the GAA, analyse it carefully, and see to what extent the GAA trends correlate with the official CSO data.

Measurement issues abound here, as David explains in his FYP, so please take this chart with several pinches of salt, but what we is a jump in CSO outward migration, and a drop in GAA player movement over the decade of study. The figure below traces out the story nicely. We see a big jump in GAA transfers over the 2003-2007 period. Any guesses why? Answers in the comments.

We also see a divergent trend for our migrating GAA players from the other CSO migrants.

The next idea was to look at monthly movement of GAA players abroad by decade in absolute terms. Guess what? McWilliams looks like he's on the money, because although the summer trend still prevails (players hitting the road for the summer), the absolute numbers of transfers hops up between 2007 and 2008, relative to the previous years, if you look at the summer movements.

David (Healy) also documents a big jump in transfers for county players from 2005 to 2008---from 0.02 of the sample in 2006 and 2007 to 0.18 of the total in 2008.

So, there's more going on here than I originally thought. I'd really like to rerun the analysis for 1980-2009/10. Does anyone out there know where I might look for GAA transfer data like that? If you do, drop me a line.