J. Ignacio García-Pérez, "Job Separation in a Non-stationary Search
Model: A Structural Estimation to Evaluate Alternative Unemployment
Insurance Systems", Journal of Applied Econometrics, Vol. 21, No. 2,
2006, pp. 245-272.

The estimation in the present paper has been made using Spanish data derived
from the Spanish Continuous Family Expenditure Survey (Encuesta Continua de
Presupuestos Familiares (ECPF)) for the period 1985-1996. The ECPF is a
rotating panel that interviews approximately 3,200 households every quarter.
One eighth of the sample is renewed quarterly and hence an individual can be
monitored for a maximum of two consecutive years. This source gives
information about unemployed workers during their spells of unemployment and
also about their post-unemployment wages, in addition to information on
income, consumption and other household characteristics.

The estimation sample consists of unemployed household heads, which is the
only group within the ECPF whose educational level is reported. I have also
restricted the sample to married men with non-working spouses in order to
reduce heterogeneity since, given the estimation procedure, I am not able to
consider many regressors.

The estimation process uses two files. gp-file1.txt contains all the
unemployment spells expanded each spell in so many observations as the
unemployment duration of this spell. Hence, we have 390 completed spells of
unemployment and 353 censored spells that are transformed into 4,375 monthly
observations. In order to compute reservation wages more quickly, we have
divided workers into groups based on a few dichotomous variables. These
groups are in an alternative dataset, gp-file2.txt that contains the same
variables used in the estimation but only for the category groups.

Both files are ASCII files in DOS format. They are zipped in the file
gp-files.zip. Unix users should use "unzip -a".

The structure of each file is the following:

gp-file1.txt

 obs:         4,375                          
 vars:           16                          
-------------------------------------------------------------------------------
variable name      variable label
-------------------------------------------------------------------------------
cod     		individual code        
depen          		dependent variable (=1 if exit from U)
dur             	unemployment duration	
cualif          	skilled worker
inem           		have unemployment benefits
paroalto     		in a period of high unemployment
joven         		age 18-30
edad45      		age 30-45
salar          		wage
w              		dummy =1 if the wage is observed
ante1992   		dummy =1 if observed before 1992
renta         		income
salante      		previous wage
salantediscreto 	discretized previous wage	
edadm45         	age 45-65
numero3        		classification number for reservation wage's c
			omputation
-------------------------------------------------------------------------------
Sorted by:  cod  dur  


gp-file2.txt

  obs:           550                          
 vars:             8                        
-------------------------------------------------------------------------------
variable name    variable label
-------------------------------------------------------------------------------
numero3    		classification number for reservation wage's
			computation
dur            		unemployment duration
cualif        		skilled worker
inem         		have unemployment benefits
ante1992    		dummy =1 if observed before 1992
salantediscreto 	discretized previous wage
edadm45         	age 45-65
paroalto               	in a period of high unemployment
-------------------------------------------------------------------------------
Sorted by:  numero3  dur