Eurostat Workshop on ISCO - 22 November 2011 Correspondence table between ISCO 88 and ISCO 08 using additional variables Eurostat Workshop on ISCO - 22 November 2011 12/10/2011
The needs To code former years of LFS and other surveys in ISCO 08 : - give the possibility to produce historical series - urgent need for ESEG-ESSnet : allow data analysis with ISCO08 with sufficient accuracy, so as to be able to choose the best ESEG prototype ; - working on european formatted files, using the same methodology, but with few variables - or every country for himself, using as much information as possible NB : du fait de l’EDP, un peu plus d’un quart des BI dans les PC (près de 30%) Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
Correspondence tables : the state of the art There is a theoretical “Correspondence table between ISCO 88 and ISCO 08” It describes all the possible combinations of ISCO 88 -4d and ISCO 08 -4d. No rule for decision Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
Some proper ways to code former years in ISCO 08 - Independent double-coding for the same individual, working with individual basic informations and using national coding programs. But problems : o basic information is not necessarily available o coding programs might fail to code all situations. Eg France : SICORE fails in 20% cases coded by an operator - or : elaborate a Correspondence table ISCO 88 ISCO 08 using additionnal variables Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
Correspondence table between ISCO 88 and ISCO 08 using additionnal variables Input : - ISCO88-3d : o european required level o No satisfactory accuracy of national files at 4-digits level - core variables : o job status : employee / self-employed o firm NACE activity : section level (NACE 1d) - non core variables o firm size (in LFS database) : 9 employees or less / 10 or more o ISCO-08 supposed required skill - level for the occupation Ouput : ISCO08-2d Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
ISCO-08 Supposed required skill - level for the occupation § 1 - primary § 2 - secondary § 3 - tertiary education 1 § 4 - tertiary education 2 eg Question in all French surveys · specialised worker · (highly)-qualified worker · employee · technician / associate professionnal · manager / professionnal · managing director · differents grades of civil servants Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
Analysis of the Correspondence table - if the variable is not available : the table points to the most likely ISCO 08 - examples where no auxiliary variable is necessary - NACE 1-d does not solve all problems : Eg : tertiary support departments inside the firm (different from tertiary ouput activities of firms) Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
Further developments - first prototype by the end of november - Validation early december - France : . national professional classification : PCS . French LFS 2010 has been coded in ISCO 88 and ISCO 08, with independent methods using PCS and additionnal variables, - other independent double-coded national LFS ? other cross-validation of the Correspondence table ? - if not, compare 2009 and 2011 structures Need for a Correspondence table ISCO 88 -2d ISCO 08 -2d ? Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011
Thank you Correspondence table between ISCO 88 and ISCO 08 using additional variables 22/11/2011