ANPE Using administrative data on jobseekers to produce labour market statistics Joëlle Chazal RESEARCH AND STATISTICS DIVISION
La publication mensuelle des indicateurs du marché du travail Données CVS LE MARCHE DU TRAVAIL EN MAI 2004 Juillet 2004 - N 28.2 Unité : milliers Mai Avril Mai Variation Variation 2003 2004 2004 sur un mois sur un an Demandes d'emploi en fin de mois (DEFM) Demandes de catégorie 1 2381,1 2430,7 2451,0 0,8% 2,9% Demandes de catégorie 1+ 6.. 2769,6 2865,5 2883,2 0,6% 4,1% Chômeurs au sens du BIT..... 2632,0 2664,0 2678,0 0,5% 1,7% Taux de chômage.... 9,7% 9,8% 9,8% 0,0 pt 0,1 pt Sources DEFM : ANPE, DARES. Source chômage BIT : INSEE. Ces statistiques concernent les demandeurs d'emploi inscrits dans les fichiers de l'agence nationale pour l'emploi. 3000 Les demandeurs d'emploi de catégorie 1 sont les personnes inscrites à l'anpe déclarant être à la recherche d'un emploi à temps plein et à durée indéterminée, ayant éventuellement exercé une activité occasionnelle ou réduite d'au plus 78 heures dans le mois. 2800 2600 L'ensemble des demandeurs d'emploi des catégories 1 et 6 recouvre toutes les personnes inscrites à l'anpe déclarant être à la recherche d'un emploi à temps plein et à durée indéterminée, y compris celles qui ont exercé une activité réduite de plus de 78 heures au cours du mois (catégorie 6). 2400 2200 Les chômeurs au sens du Bureau international du travail (BIT) sont les personnes sans emploi, à la recherche effective d'un emploi et immédiatement disponibles. Ces données sont estimées mensuellement à l'aide d'un modèle économétrique, calé une fois par an sur les résultats de l'enquête emploi de l'insee. 2000 1800 mai-00 nov-00 mai-01 DEFM CVS de catégories 1 et 6, en milliers Chômeurs B.I.T. données CVS, en milliers DEFM CVS de catégorie 1, en milliers nov-01 mai-02 nov-02 mai-03 nov-03 mai-04 2
ANPE Jobseekers are registered on an official list managed by France s National Employment Agency (ANPE) The Agency delegates some of its tasks relating to management of the list to the unemployment insurance fund (UNEDIC)
ANPE Jobseeker registration and counting procedures
Registering jobseekers Two-step process: 1 The jobseeker goes to the unemployment insurance office (ASSEDIC): - to register as a jobseeker - to apply for unemployment benefits 2 The jobseeker goes to the National Employment Agency (ANPE): for a preliminary job interview and to prepare for job hunting 5
Registration Registration takes place at the unemployment insurance office (ASSEDIC) - Identification - Reasons for registering as a jobseeker - Type of job sought: * long-term, seasonal, full-time, part-time - Availability: * immediate or deferred availability: training, sick or maternity leave, in work + work certificates required to apply for benefits 6
Preliminary job interview at the National Employment Agency (ANPE) This interview is madatory It takes place at the local ANPE office Its purpose: - recording helpful information for job placement: * skills, types of jobs sought - explaining the Agency s job hunting assistance services 7
Jobseeker categories Categories are determined by combining 3 criteria: - Availability for work: immediate or deferred - The type of job sought: long-term or temporary, full-time or part-time - Working at casual or occasional jobs more than 78 hours per month The combinations produce a total of 8 jobseeker categories 8
Jobseeker categories Jobseeker categories IMMEDIATE AVAILABILITY FOR WORK Working at occasional jobs more than 78 hours per month Type of job sought: no yes - Full-time permanent job Cat 1 Cat 6 2,307,000 455,000 - Part-time permanent job Cat 2 Cat 7 431,000 81,000 - Temporary job Cat 3 Cat 8 257,000 111,000 NOT IMMEDIATELY AVAILABLE FOR WORK Reason for unavailability Sick leave or maternity leave Cat 4: 178,000 In work, in training Cat 5: 104,000 Data: May 2004 9
Jobseeker categories Jobseeker categories Jobseeker count by categories 4500000 4000000 categories 1 2 3 6 7 and 8 3500000 3000000 categories 1 and 6 2500000 2000000 categorie 1 1500000 janv-96 janv-97 janv-98 janv-99 janv-00 janv-01 janv-02 janv-03 janv-04 10
Monthly renewal of registration Jobseekers must renew their registration at the end of each month: - by letter, telephone or over the Internet At the same time, they must report occasional work done during the month and - Case No. 1: renew their registration - Case No. 2: not renew their registration (they are required to report the reason for non-renewal: starting a job or training, etc.) Jobseekers who fail to renew their registration are no longer deemed to be available for work 11
Consequences of non-renewal of registration Jobseekers who fail to renew their registration are no longer deemed to be seeking work and they are taken off of the list for: failure to sign on * This description covers different situations, such as: - Jobseekers who forget to renew and register again the next month => leads to overestimation of jobseekers exiting unemployment - Unreported job placings => jobseekers taken off of the list in this case are not counted as exiting unemployment 12
Data What administrative data tell us about jobseekers: Characteristics of the jobseekers: sex, age, skills, educational attainment, etc. * Characteristics of the job hunt: start date, end date, status: ongoing or completed, category of job sought, reasons for registering or exiting unemployment (when reported), etc. The data are based on a complete count 13
ANPE Production chain for monthly statistics
Statistical production chain: data for month m ANPE ASSEDIC Data entered local office local office on local computers Data feed ASSEDIC computer centre transfer 20th of month (m+1) ANPE computer centre transfer 25th of month (m+1) Ministry: DARES INSEE Management of joint ANPE/Unédic database reception and processing production audits: volume and plausibility analysis and publication end of month (m+1) estimate of ILO unemployment 15
Statistical chain: data for month m La chaîne statistique : données du mois m Stocks data: end of month m-1 Stocks data: end of month m Exiting JSs New JSs Account balance: Stock m = stock (m-1) + New JSs - Exiting JSs Plausibility audits by region and category 16
Limitations of this data source * Administrative data source derived from the list of jobseekers used for statistical purposes: - affected by changes in the law and in list management methods * Based on the jobseekers answers: - 30% of the jobseekers entering or exiting the ANPE list do not report their reasons for doing so 17
Limitations of this data source (continued) * Month-by-month picture, with no tracking of individuals over time: - problems measuring the recurrence of unemployment - problems measuring long-term unemployment 18
Overcoming limitations and shortcomings Historical database on jobseekers Complementary surveys (examples: Jobseeker Tracking Survey, survey of jobseekers taken off of the ANPE list, etc.) 19
ANPE Historical database: Longitudinal tracking of jobseekers
Historical database Jobseekers tracked for at least 10 years: Audited and adjusted compilation of monthly "snapshots" (adjustment example: elimination of jobseekers exiting the list in the last three days) 1 complete database and 1 sample database These are used to: - track cohorts - calculate cumulative unemployment spells - analyse the recurrence of unemployment - measure the speed of exiting unemployment 21
Examples: tracking new jobseekers in 2000 over 36 months 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1 2 3 4 5 6 7 8 9 101112131415161718192021222324252627282930313233343536 S ituation 12 months after regis tration - off list at end of month - 1st exit 50% - off list at end of month - but having exited and registered again 9% - on list at end of month - but having exited previously 14% - on list continuously since month of registration 26% 22
Examples: Other approaches for long-term unemployment: 12 months of unemployment Jobseekers at end of 2003 - all categories 1 933 1 339 651 registered for 12 mo. s registered for 12 mo. s wit h no occasional work registered for 12 mo.s during last 18 mo. s 23
Examples: Measuring the recurrence of unemployment: -35% of the 1,250,000 new jobseekers in the 4 th quarter of 2003 had already been registered in the previous 6 months Measuring job placings: - 24% of the 3,862,000 jobseekers registered at the end of December 2003 exited the list in the first quarter of 2004-75% of new jobseekers in the 4 th quarter of 2002 exited the list at least once in the 12 months following their registration 24
The measurements can be made - for different types of populations (young people, women, disabled workers, etc.) -for different geographical units (Regions, Départements, etc.) - Limitations of the historical database: It only contains data on unemployment spells (no data between unemployment spells) 25