Wurtz 19
Poland reliberalized by 2001, while Slovakia, Columbia, and Guinea had not. Due to data
availability with some of the economic variables, the sample size falls to 95 countries.
Data analysis
In this paper I use event history analysis to test the first set of hypotheses concerning the
timing of loosening restrictions on FC deposits. Event history, or duration, analysis is useful for
analyzing the decision to allow FC deposits because it measures the probability of switching
from one state of the world to another for a given period of time. Even history studies allow us
to analyze how long a particular event occurs in a group of units, as compared to a control group.
In this case, I am studying how long restrictions on FC deposits occur. By comparing those
countries that receive the “treatment” to those that do not, I can determine the probability for a
given year that a country would allow FC deposits.
As the world moved from the early Bretton Woods system of capital immobility and
capital to controls to the present day, governments have relaxed their regulations at different
times. The time series nature of event history analysis allows us to test the timing of reform
programs while controlling for country changing economic conditions and other country specific
variables over time. It also allows us to test for non-varying covariates such as veto players,
democracy, or other institutional variables.
My data begin in 1983, thus the 37 countries that already allowed FC deposits as of 1983
do not appear in the data set. I also do not include countries that enter the IMF Annual report
having already allowed FC deposits. This reduces the number of available cases and introduces
some amount of bias. However, including these cases would also introduce its own form of bias
Convention