EXTENSIVE VS. INTENSIVE MARGIN: CHANGING PERSPECTIVE ON THE EMPLOYMENT RATE. and Eliana Viviano (Bank of Italy)

Transcription

1 EXTENSIVE VS. INTENSIVE MARGIN: CHANGING PERSPECTIVE ON THE EMPLOYMENT RATE Andrea Brandoln and Elana Vvano (Bank of Italy) 2 European User Conference for EU-LFS and EU-SILC, Mannhem 31 March 1 Aprl, 2011

2 Motvaton (1): Comparsons of European labour markets are typcally based on employment rates. For nstance: The Lsbon Strategy: Target: 70% employment rate (15-64) by 2010; Europe 2020: Target 75% (20-64) by 2020 Household level ndces of non-employment whch mmc the employment rate (EU2020 poverty target partly based on jobless households) However: 1. The employment rate s based on a yes/no decson to work (people workng 1 hour and 40 hours treated n the same way) 2. European labour markets dffer because of nsttutons 1. In 2007 the share of part-tme workers on total employment was equal to 26 per cent n Germany, to 14 n Italy and to 47 percent n the Netherlands. 2. The share of fxed-term employees on total employees was close to 32 percent n Span, to 15 percent n Germany and France, 13 percent n Italy, 28 percent n Poland, 22 n Portugal. How to take nto account all these aspects?

3 Motvaton (2): Necessary to pay attenton also to How much people/households work (ntensve margn or work ntensty) Today, there s ncreasng attenton on work ntensty, defned as number of months worked /12 months per year (share of ndvdual/households wth very low work ntensty), amed at takng nto account closed-end contracts Another dmenson s totally mssng: hours worked per week/day/year? Why? Probably because of measurement ssues Tradtonal vew: hours worked per week/day are determned by labour supply. Part tme s an ndvdual choce to reconcle work and lesure (chldcare ) But s t stll true? In some sectors part tme jobs are very wdespread A szeable share of part tme n Europe s nvoluntary Durng 2009 a rapd ncrease n nvoluntary part tme

4 Objectve: to calculate a weghted employment rate, whch accounts for the ntensve margn (also work ntensty ). A unque ndex. Our paper: The ntensve margn of labour s the combnaton of: hours worked per week by ndvduals (to account for dfferences n full/part tme employment), over total workable hours; months worked per year by ndvduals (to account for dfferences n lmted duraton employment as n Laeken ndcators and Europe 2020). Can be calculated at the household level It s more closely related to socal excluson

5 Data: EU-SILC Wave 2007 for EU24 + Iceland + Norway Why EU-SILC? EU-SILC only has nformaton on months worked per year (not n EU-LFS) EU-SILC has household varables (not avalable yet n EU-LFS) Unfortunately EU-SILC defnton of employment s not fully comparable wth EU-LFS for ALL countres.

6 The standard ILO employment rate It deals wth extensve margn only: to work or not. People who work just 1 hour per week and people workng 40 hours per week are equally treated. E = P I I I = = 1 0 f f employed not empl. P = workng age populaton

7 Fg: 1 Iceland Denmark Norw ay Netherlands Sw eden Unted Kngdom Austra Cyprus Fnland Estona Germany Ireland Latva Slovena Portugal Czech Republc Span Lthuana France Luxembourg Belgum Greece Slovaka Italy Hungary Poland Standard result: low employment rate n Southern Europe

8 Weghted employment rate by work ntensty (hours+months) Eˆ = ω P I ω = no. hours EU medan per week hours (40h) no. months 12 work months per year Hours work nt. Months work nt. Weghted employment rate, equal to standard employment rate f all workers have a full tme (40 hours) permanent jobs.

9 Iceland Sw eden Latva Norw ay Unted Kngdom Estona Czech Republc Lthuana Cyprus Slovaka Denmark Luxembourg Greece Fnland Span Portugal Austra Slovena Germany Poland Hungary France Ireland Italy Belgum Netherlands 11th n FIG. 1 12th n FIG. 1 4th n FIG. 1 2nd n FIG. 1 22th n FIG. 1 Fg: 2 Rank changes consderably: Lower employment rates n Netherlands, Ireland, Germany and Denmark Hgher rates n Greece and Span, Poland and Hungary

10 Exercse 1: Aggregaton by household Easy to aggregate at the household level: Snce we have a contnuous measure of employment, t s easy to aggregate t at the household level. 1. sum employment work ntensty of all household members 2. dvde total household employment by equvalzed household sze 3. assgn each ndvdual the average (equvalzed) per capta work ntensty Ths s an ntensve margn verson of the jobless household rate ~ E = 1 P W W = 1 s h h ω I h = household of the -th ndvdual Sh= equvalzed household sze of the -th ndvdual

11 Fg. 3 Iceland Cyprus Latva Estona Lthuana Sw eden Czech Republc Slovaka Luxembourg Norw ay Slovena Greece Span Portugal Unted Kngdom Fnland Austra Italy Denmark Poland France Hungary Germany Belgum Ireland Netherlands 2nd n FIG. 1 11th n FIG. 1 12th n FIG. 1 4th n FIG. 1 22th n FIG. 1 6th n FIG.1 Per capta work ntensty: household level

12 Exercse 2: Decomposton: what dmenson of ntensty matters? Months or hours? Country W. Empl. rate 100-empl. rate Effect p.t. Effect fx. Term c. Non-employment Austra Belgum Germany Denmark Span France Greece Hungary Ireland Italy Netherlands Poland Portugal Sweden Europe (26 countres)

13 Exercse 3: Inequalty n the dstrbuton of employment Recent lterature focuses on nequalty n the dstrbuton of work (e.g. Gregg and Wadsworth, 2008, 2010). Are there households whch are excluded from employment? Is employment concentrated among relatvely few households? Work ntensty helps to answer ths queston by usng standard tools for studyng nequalty E. g.: Gn ndex of the dstrbuton of work ntensty of European households (ndces allowng for decomposton by socodemographc groups may be used too)

14 Exercse 3: Inequalty n the dstrbuton of employment Country Hours Month Overall Austra Belgum Cyprus Czech Republc Germany Denmark Estona Span Fnland France Greece Hungary Ireland Iceland Italy Lthuana Luxembourg Latva Netherlands Norway Poland Portugal Sweden Slovena Slovaka Unted Kngdom Europe (26 countres)

15 Exercse 4: Dsposable ncome and work ntensty Is work ntensty nformatve about poverty? Partal correlaton of month and hour work ntensty wth household dsposable ncome Month work ntensty Hours work ntensty Austra Belgum Germany Denmark Span Fnland France Greece Hungary Ireland Italy Netherlands Poland Portugal Unted Kngdom

16 Conclusons Summng up: People choose not only whether to work, but also how much to work. Ths choce (ntensve margn) s nfluenced by nsttutons and structural characterstcs (avalablty of part tme jobs, lmted duraton contracts, etc.). Depends also on the wthn-household allocaton of work (t matter especally for women). Two man dmensons of work ntensty: () hours per week; () months per year. Takng nto account these aspects n a comprehensve framework by usng a weghted employment rate, we fnd that n France, Germany, Italy and UK employment rates are rather close. Our fndngs suggest that: Cross-country comparsons should be carred out by the use of flexble ndces lke the ones proposed here, whch allow to control for large nsttutonal and cultural dfferences. Need of comparable and hgh-qualty nformaton of employment of ndvduals durng the year; hgh-qualty nformaton on households. The relatonshp between employment status and poverty can be analyzed n a more flexble framework than the standard extensve margn framework.