search

ijyliu / ECMA-31330-Project

Econometrics and Machine Learning Group Project
2 stars 1 forks source link

Revolutions in full democracy #6

Closed nicomarto closed 3 years ago

nicomarto commented 3 years ago

The idea is to simulate Acemoglu's Model for revolution in full democracies. We could potentially extend it to more political regimes.

The model can be found in Chapter 19 of Acemoglu's political economy lecture notes: https://raw.githubusercontent.com/ijyliu/ECMA-31330-Project/main/Political%20Economy%20Lecture%20Notes.pdf?token=AQCGE4MKSRR2STW4Q6QHAH3ARCXEG

After simulating, we could go and gather data for a panel dataset containing voters turnout and political conflict measures, and see how it fits the data

ijyliu commented 3 years ago

For turnout- consider a nonlinear relationship?

Maybe we construct the index, and then see how it performs in terms of explanation.

Maybe voter turnout is the variable, but maybe it's something else. Representation index of voter turnout, but also a bunch of other items. Political system, etc. variables.

There are also a lot of political instability measures to use.

ijyliu commented 3 years ago

Check the data, specifically dependent variables section of my paper here: https://raw.githubusercontent.com/ijyliu/Commitment_Institutions_Instability/master/Release/Paper/Liu-%20Commitment%20Institutions%20and%20Instability.pdf

Here's the vdem codebook, where you can see the massive number of variables it contains: https://www.v-dem.net/en/data/reference-material-v11/ . Lots of variables + relatively few country/time periods in the panel = good for the dimensionality reductions we've been doing in this class.

There's also this dataset, which I didn't end up using for my thesis: https://sites.google.com/a/nyu.edu/adam-przeworski/home/data

ijyliu commented 3 years ago

Check course notes Chapter 5 and https://dspace.mit.edu/bitstream/handle/1721.1/111106/Grouped-patterns-Bonhomme_Manresa-Quirante.pdf?sequence=1&isAllowed=y