J. Paul Elhorst, Pim Heijnen, Anna Samarina, and Jan P. A. M. Jacobs,
"Transitions at Different Moments in Time: A Spatial Probit Approach",
Journal of Applied Econometrics, Vol. 32, No. 2, 2017, pp. 422-439.

DATA FILES 

There are three estimation data files and four spatial weights
matrices files, all in plain ASCII (text) format. They are zipped in
ehsj-data.zip. Unix/Linux users should use "unzip -a".

Data files are:

  Y (country, year, adoption) -- includes countries with adoption dates
  for IT.

  X_raw is the data file for explanatory variables, as retrieved raw
  from data sources.

  X_imputed is the file with explanatory variables imputed to get a
  balanced panel dataset.

Spatial weights matrices include:

  W (10 nearest neighbours matrix),
  A (common language matrix),
  B (common legal origin matrix), and
  C (BIS membership matrix).


DATA SOURCES

1. Classification of countries as Inflation targeters and
   non-inflation targeters. Sources: central banks documents and
   literature (see references in the paper)

2. Inflation - Annual CPI inflation rate, transformed. Sources: IFS
   IMF, Datastream.

3. Output growth - Annual GDP growth rates, in %. Sources: IFS IMF.

4. Exchange rate regime Indicator, from 1 (hard peg) to 4 (freely
   floating). Source: Reinhart and Rogoff(2004), Ilzetzki et al. (2011).

5. Government debt Central government debt as % of GDP. Sources:
   Datastream, OECD Stat, Jaimovich and Panizza (2010).

6. Financial development - Domestic credit provided by the financial
   sector/GDP. Sources: WDI World Bank.

7. Central bank instrument independence - Dummy variable, 1 - central
   bank is instrument independent, 0 - otherwise. Sources: Cukierman et
   al. (2002), Arnone et al.(2007), central banks laws.


VARIABLE NAMES

Year	1985-2012
IT	1 = a country adopted Inflation Targeting in year t, 0 = otherwise
INFLT	CPI inflation rate
GDPG	annual GDP growth rate
EXRR	Exchange rate regime indicator 
DEBT	Central government debt as % of GDP
FINDEV	Financial development - domestic credit to GDP
LCBI	Central bank instrument independence, 1 = central bank is instrument
        independent, 0 = otherwise
W	Binary contiguity matrix, for 10 nearest neighbours
A	Common language weights matrix
B	Common legal origin weights matrix
C	Common BIS membership weights matrix

NUMBER OF OBSERVATIONS: 1628 (for 1985-2012), 1392 (for the analysed
period, 1985-2008)


MATLAB FILES

Four zipped folders provide Matlab programs used to estimate the
results in Tables 4, 5, and A.1. (Appendix) in the paper.

standard-spatial-probit.zip contains a Matlab folder to reproduce the
estimation results in Table 4, column 1; our-spatial-probit.zip is for
Table 4, column 2.

The programs in marginal-effects.zip (rin_regressions_effects)
produces the results in Table 5. Note, however, that the results will
never be exactly the same, since these effects (as explained in the
paper) are obtained by randomly drawing from the variance-covariance
matrix 100 times. These random draws are different for every new run.
The programs in our-spatial-probit-time-dummies.zip reproduce the
results in Table A.1.