Econometrics workshop

[Mikolaj Malinowski]

Dear all,

Regarding our workshop coming Thursday. I have constructed a datafile which I wanted to use to investigate a glass celling for development of medium cities in Poland. The datafile shows that there was no celling so I am at square one again. I know that the material is nice, but, as for now, I lack the idea what I can do with it now to make it an interesting story (still dreaming about real wages paper for Pim). The theme is low urbanisation or what drives urbanisation. Please just look at the datafile for five minutes and tell me what kind of sexy empirical strategy I can take to digest it or perhaps it answers some important questions you once had.

There are only few points in time therefore all my experience with time series analysis is worth little.

My ideas  for now:
1)probit/tobit model which explains/shows what is so special about the successful cities
2)Cities were levelled out in 1650, some of them recovered others did not. A nice regression of the levels/fd on all the dummies, without the presence of the infamous endogeneity problem.
3)the original idea was that the international trade boosted development of the cities with market access that later overshadowed the rest, perhaps it can still be shown/was the case. Would hate to add GPS to it to make use of gravity model though, too many cities. Any idea how to show it?

Need to pick up your brains for brilliant ideas.

Mikolaj
Ps. do not circulate this.

[Foldvari, P. (Peter)]

Dear Mikolaj,

I understand if you really want to avoid using GIS, but essentially for this type of research that is a quite standard methodology. You can alternatively transform population or other size measures into a useable dependent variable and make a panel and add hunderds of dummies, that would even be enough for the PhD, but that will not lead to any revolutionary good paper.
Also, what is your theory? Looking for a methodology without a clear theory is like trying to find a pin in a haystack.

Best ,
Péter
_______________________________________
From: econometrics...@googlegroups.com [econometrics...@googlegroups.com] on behalf of Mikolaj Malinowski [mikolaj.m...@gmail.com]
Sent: Thursday, October 04, 2012 4:30 PM
To: econometrics...@googlegroups.com; Piotr Guzowski
Subject: Econometrics workshop

--
You received this message because you are subscribed to the Google Groups "Econometrics Discussion Group ESH" group.
To post to this group, send email to econometrics...@googlegroups.com.
To unsubscribe from this group, send email to econometricsdiscus...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Mikolaj Malinowski]

Dear Peter,
Thanks for the suggestions. It seems that i have to add the possition to do something interesting. Perhaps i could introduce ranks instead of population. The Krugman (NEG) ideas always seemed appealing to me. I think investigation of domination of aglomerations and links between agglomeration formation and transportation costs are also interesting.

Mikolaj Malinowski