Bart Cockx and Muriel Dejemeppe, "Duration Dependence in the Exit Rate out
of Unemployment in Belgium. Is it True or Spurious?", Journal of Applied
Econometrics, Vol. 20, No. 1, 2005, pp. 1-23.

Our analysis exploits quarterly census data relative to the unemployed male
population in Wallonia (Belgium). These data are the property of the 
« Office National de l'Emploi » (ONEm) and the « Office communautaire et
Régional de la Formation Professionnelle et de l'Emploi » (FOREM), Belgium.

The data are stratified by the quarter of inflow into unemployment, the
unemployment duration and the age group (<= 28 years old, and 29-44 years
old). The inflow in a given quarter is equal to the number of laid-off
workers who are still unemployed at the end of the quarter in which they
enter (an exit should also last at least three months to be recorded as 
such). The observation period consists of 19 quarterly intervals, the first
interval starting on the 1st of June 1989, the last one ending on the 28th
of February 1994. So there are 18 cohorts of unemployed (a cohort being
defined by its quarter of inflow), stratified by age group (there are 18
cohorts instead of 19 as there is no exit defined for entrants in the 19th 
quarter). For each cohort, the flows out of unemployment are counted on a
quarterly basis, from the time of inflow until the end of February 1994, the
date at which all spells are right censored. For each age group, the data
are therefore grouped into 171 (=1+2+...+18) homogeneous cells, from which
probabilities of leaving unemployment are computed. 

There are therefore two data sets: Oldmen.dat (Men aged 29-44) and
Youngmen.dat (Men aged <=28), with 171 observations each. Both these files,
which are in DOS format, are zipped in the file cd-data.zip

In each data set, there are 11 variables described below:

ANE (89,90,...,93), character: Year of entry into unemployment

TRIME (1,2,3,4), character: Quarter of entry from unemployment 
Remark:
1=(Mar, Apr, May)
2=(Jun, Jul, Aug)
3=(Sep, Oct, Nov)
4=(Dec, Jan, Feb)

ENTRY (0,2,...,17), numeric: Calendar time at entry (in quarters)
Remark: l in the paper

ANI (89,90,...,94), character: Year of exit from unemployment

TRIMI (1,2,3,4), character: Quarter of exit from unemployment
Remark:
1=(Mar, Apr, May)
2=(Jun, Jul, Aug)
3=(Sep, Oct, Nov)
4=(Dec, Jan, Feb)

EXIT (1,2,...,18), numeric: Calendar time at exit (in quarters)
Remark: l+k in the paper

DUT (1,2,...,18), numeric: Unemployment duration (in quarters)
Remark: k in the paper

SCOUNT, numeric: Number of individuals who entered unemployment at 'ENTRY'
and are still unemployed at the end of the duration quarter 'DUT' Remark:
uk+1l in the paper

SLCOUNT, numeric: Number of individuals who entered unemployment at 'ENTRY'
and are at risk of leaving unemployment within the duration quarter 'DUT'
Remark: ukl in the paper

PROBA, numeric: Probability of leaving unemployment
Remark: Pkl in the paper, =(SLCOUNT-SCOUNT)/SLCOUNT

LENTREED, numeric: De-seasonalised logarithm of the number of quarterly
flows into unemployment (computed from the data set) Remark: b(l) in the
paper


Please address any questions to:

Muriel Dejemeppe

IRES
Place Montesquieu 3
B-1348 Louvain-la-Neuve
Belgium

dejemeppe@ires.ucl.ac.be