Analytical support in the setting of EU employment rate targets for Working Paper 1/2012 João Medeiros & Paul Minty

Transcription

1 Analytical suort in the setting of EU emloyment rate targets for 2020 Working Paer 1/2012 João Medeiros & Paul Minty

2 DISCLAIMER Working Paers are written by the Staff of the Directorate-General for Emloyment, Social Affairs & Inclusion, or by exerts working in association with them. Views exressed do not necessarily reflect the official views of the Euroean Commission. Comments and enquiries should be addressed to the: Euroean Commission Directorate-General for Emloyment, Social Affairs & Inclusion Unit A2 B 1049 Brussels, Belgium Euroean Union Cover hotos: 123RF Legal notice: Neither the Euroean Commission nor any erson acting on behalf of the Commission may be held resonsible for the use that may be made of the information contained in this ublication. For any use or reroduction of hotos which are not under Euroean Union coyright, ermission must be sought directly from the coyright holder(s). May 2012 I 2

3 Table of Contents Abstract Introduction Historical background The 2010 emloyment rate targets ER develoments since 2000 and outlook to Progress between 2000 and Develoments over 2009 a shar setback Outlook to Format for the 2020 target The imact of rising levels of education The age ranges for the ER target Develoments in the age rofile of emloyment rates The uer age limit The lower age limit The ractical effect of adjusting age bands on emloyment rates and coverage of total emloyment Possible values for the 2020 target Aroach Eurostat's 2008 baseline scenario for demograhic variables The methodology for rojecting labour market variables Particiation rates Unemloyment rate assumtions for Emloyment Rate rojections for Considerations in setting otential ER targets for Criteria for setting ambitious yet feasible targets Further considerations for setting targets Exloring the imlications of setting articular targets Original target oulation Revised target oulation Older workers target oulation Comarison of required rogress to meet targets from the labour market rojections Proosed common ER targets for Setting comlementary and differentiated targets - otions for setting national emloyment rate targets for Rationale for comlementary differentiated targets Possible aroaches to setting national ER targets for Otions examined Reference emloyment rates Results of otions examined Initial examination of results Statistical comarison of the otions Better outcomes otions' ER targets versus historical values and LM rojections for the reference oulation Comarisons in terms of emloyment growth Overall assessment of the three 'better outcomes' otions for setting national targets Summary Eilogue Annex I Annex II Annex III May 2012 I 3

4 List of Tables Table 1: Develoments in overall, female and older workers' emloyment rates, Table 2: Trends in emloyment rates by educational attainment ( ) Table 3: Particiation and emloyment rates in 2007 in the EU Table 4: Effects of using adjusted age bands for the working age oulation, EU27 in Table 5: Working age oulation in the EU27 by age bands (millions) Table 6: Particiation rates Table 7: Unemloyment rate assumtions for 2020 (age band 15-64) Table 8: Emloyment rates in the EU27 for the working age oulation Table 9: Emloyment rates in the EU27 for women Table 10: Emloyment rates in the EU27 for men Table 11: Total oulation aged Table 12: Female oulation aged Table 13: Total oulation aged Table 14: Female oulation aged Table 15: Older workers oulation aged Table 16: Efforts required to reach round ER targets in Table 17: Emloyment rates (age grou 20-64) from rojection scenarios and the best historical rate in the eriod Table 18: ER imlications of various otions for distributing the required effort to reach an EU average ER (20-64) target of 75% in Table 19: Imlied ER rises on 2010 values under the various otions Table 20: Statistical analysis of the ER results from the otions Table 21: Target ERs under the referred otions coomared to reference ERs from historical erformance and rojection scenarios Table I. 1: Ages of entitlement to standard and early retirement ensions across EU Member States, Table I. 2: Comulsory education in Euroe 2007/ Table II. 1: 'Otimal' arameters for various reference values Table II. 2: Actual arameterisation of some Otions Table III. 1: Actual arameterisation of Quadratic Otions List of Charts Chart 1: Develoments in overall, female and older workers' emloyment rates in the EU Chart 2: Forecast emloyment growth rates over 2009 and Chart 3: Forecast change in ERs across EU Member States between 2008 and Chart 4: Age rofile of emloyment rates in the EU27 in 2000 and 2007, and 2020 rojections Chart 5: Emloyment rates in the EU27 by individual year of age, 2000 and Chart 6: Demograhic rofile of EU27 oulation aged u to 79 in 2000, 2010 and Chart 7: Particiation/enrolment in education in the EU27 by age in 2000 and Chart 8: WAP in the EU27 (millions) Chart 9: Age distribution of the working age oulation (15-64) Chart 10: Labour market rojections (the main stes in colour/darker) Chart 11: Evolution in overall emloyment rates (15-64) across EU Member States, Chart 12: Different Otions shift the effort across MS Chart 13: Otions involving equal emloyment growth (4 and 4a) May 2012 I 4

5 Chart 14: Better Overall Outcomes Otions Chart 15: Otion 1a - ER targets (20-64) and reference values Chart 16: Otion 3 - ER targets (20-64) and reference values Chart 17: Otion 3a - ER targets (20-64) and reference values Chart 18: Comarison of required ER rises and average emloyment growth rates Chart 19: Average emloyment growth rates in to achieve the ER targets under otions 1a, 3 and 3a Chart 20: Annual emloyment growth rate required to meet targets under otion 1a comared to that under the rojection scenarios Chart 21: Annual emloyment growth rates required to meet targets under otion 3a comared to that under the rojection scenarios Chart 22: National emloyment rates in 2011 and targets set by Member States in their 2011 National Reform Programmes May 2012 I 5

6 Abstract This aer rovides an overview of the analysis carried out during 2009 and 2010 with regard to the setting of a new overall emloyment rate target for 2020 as art of the Euroe 2020 strategy, including the reference oulation to be covered as well as the aroriate value for the target. As such it rovided analytical suort leading to the decision of the March 2010 Euroean Council to set an overall emloyment target for the EU27 in 2020 of 75% for those aged It also summarises the subsequent analysis carried out to investigate ossible aroaches to translate the common target into secific targets for individual Member States, the results of which rovided analytical suort to the ensuing dialogue between individual Member States and the Commission in this context during mid The aer is essentially methodological and is mainly intended as a historical record of the aroach used and considerations taken into account when roviding analytical suort to the discussions on the Euroe 2020 emloyment target. As such it should be a useful reference of assistance in similar future target setting exercises. May 2012 I 6

7 1. Introduction This aer rovides an overview of the analysis carried out in DG Emloyment, Social Affairs and Equal Oortunities during 2009 and 2010 with regard to roviding analytical suort to the setting of a new overall emloyment rate (ER) target for 2020 as art of the Euroe 2020 strategy, following on to that set for 2010 as art of the Lisbon Strategy. The aer draws substantially from documents reared for discussions in the Emloyment Committee and its indicators sub-grou in rearation of the setting of the Euroe 2020 emloyment rate target, and the authors gratefully acknowledge the valuable feedback rovided by these grous. This aer is ublished now as the first in a newly launched series of technical working aers. It is mainly methodological and is intended as a historical record of the aroach used, based on the data available at the time, and the considerations taken into account when roviding analytical suort to the discussions on the Euroe 2020 emloyment target, and as such should be of assistance in similar future target setting exercises (such as any mid-term review of the Euroe 2020 strategy). It rovides value as a record of the key considerations then revalent and to illustrate the debate which occurred at the time. The focus of the aer is on the emloyment rate, i.e the caacity of an economy to rovide work (emloyment) for all those who can be considered as wanting to work (in the working age oulation). The document rovides first the contextual setting for emloyment rate develoments over the revious decade and the initial imact of the recession, which by lowering ER values at the beginning of the new decade conditioned the setting of a new target for Following this there is a summary of the review carried out on the format of the ER target, in articular the issue of the aroriate age range to be covered in the calculation of the new headline emloyment rate. Several develoments, including EU objectives for furthering educational attainment, articularly for youth, and extending working lives for older workers, have led to concerns over whether the age range (15-64) used for the overall Lisbon emloyment rate target remained aroriate. Given the objectives for raising education levels, lowering school droout rates and taking advantage of higher longevity and imrovements in health conditions to extend labour market articiation among older workers, the merit of adjusting the current age brackets was investigated, eventually leading to the decision to focus on the new target oulation aged The following section gives an overview of the analysis on the ossible values this new overall target might take, based on demograhic rojections and economic forecasts available at the time, which took into account not only anticiated future trends and challenges but also the ossible negative effects on otential outut and emloyment arising from the severity and duration of the recession. Finally, a summary of the technical analysis carried out on why and how the overall target might ossibly be translated into otential national emloyment rate targets for Member States is resented Historical background 2.1 The 2010 emloyment rate targets Achieving higher emloyment rates (ERs) has been at the core of the Lisbon Strategy, the main strategy at EU level for growth and jobs over the first decade of this century. Under the Lisbon Strategy the EU set itself three emloyment rate targets to be reached 1 The information set used includes annual LFS data for 2008 and the autumn 2009 Euroean Commission economic forecast. May 2012 I 7

8 by 2010: i) an overall target of 70%; ii) a 60% target for women; and iii) a 50% target for older workers 2. The main rationale for these targets has been to suort job creation and economic growth. Furthermore, setting targets is considered to be a good way to kee the focus on the most imortant olitical challenges, contributing to evidence-based olicy making. The targets on the overall emloyment rate and the female emloyment rate were first set in Lisbon, in March The target on the emloyment rate for older workers was set in Stockholm, in March Policies should contribute to achieving an average emloyment rate for the Euroean Union (EU) of 70 % overall, of at least 60 % for women and of 50 % for older workers (55 to 64) by The main indicators to monitor these targets are the Emloyment rates (total oulation aged 15-64, women aged 15-64) measuring the ersons in emloyment as a roortion of the total oulation in the same age grou (data source: EU labour force survey (LFS)). The Emloyment rate for older workers measures the ersons in emloyment in the age grou as a roortion of the total oulation in the same age grou (data source: EU LFS). 2.2 ER develoments since 2000 and outlook to Progress between 2000 and 2008 Since the beginning of the Lisbon Strategy, considerable rogress had been made in EU labour markets before the crisis in financial markets deeened sharly in autumn 2008 and led to economic recession - the overall emloyment rate had risen by close to 4 ercentage oints (s), reaching 65.9% in 2008 (comared to 62.2% in 2000), reflecting a rise in the total number of eole in emloyment 3 of around 19 million. The ERs for women and older workers in articular showed considerable rogress, reaching 59.1% and 45.6% resectively in 2008 (Chart 1). Chart 1: Develoments in overall, female and older workers' emloyment rates in the EU Overall ER (%) Female ER (%) Older workers ER (%) 60 % of resective age oulation Source: Eurostat, EU Labour Force Survey Year Comared to 2000, by 2008 the overall emloyment rate in the EU27 had risen 3.7 ercentage oints, the female emloyment rate 5.4 s, and that for older workers by 2 Peole aged between 55 and National concet, based on Eurostat LFS data May 2012 I 8

9 8.7 s (Table 1). As a result, the gas relative to the targets were resectively 4.1, 0.9 and 4.4 s in However, one should bear in mind that the targets were originally set when the EU consisted of just 15 Member States, and with regard to that comosition rogress towards the targets has been more substantial, to the extent that the gas in 2008 were a more limited 2.7, and 2.6 s resectively. As a result, the mechanical imact of the two recent enlargements has been to reduce average EU emloyment rates by around ercentage oints, deending on the secific target. Table 1: Develoments in overall, female and older workers' emloyment rates, EU27 EU25 EU Change Change Change Overall ER (%) Female ER (%) Older workers ER (%) Source: Eurostat, EU Labour Force Survey Develoments over 2009 a shar setback In contrast to the eriod of strong labour market erformance u to mid-2008, labour markets exerienced a shar setback over The unrecedented crisis in global financial markets which gathered ace in autumn 2008 led to the most severe recession since the Great Deression, affecting the wider economy and increasingly imacting on labour markets in the EU. After several years of relatively strong labour market erformance, the economic crisis led to a sudden reversal of emloyment growth which threatened to swee away much of the success of recent years. By mid-2009 emloyment in the EU had contracted by 4.3 million (1.9%) comared with a year earlier, resulting in a dro in emloyment rates, while unemloyment rose strongly. Labour market erformance worsened across all Member States, but most notably in the Baltic States, Ireland and Sain. In contrast, in other Member States such as Belgium and Germany, emloyment showed remarkable resilience to the crisis, in art due to extensive recourse to short-time working arrangements and other (temorary) measures. Indeed, several countries (esecially AT, BE, DE, IT and NL) have steed-u measures to kee workers in emloyment by adjusting roduction and working time. Wage moderation and a higher incidence of wage concessions have also heled maintain jobs in several sectors Outlook to 2011 According to the autumn 2009 Commission forecast available at the time of analysis, the EU economy was set to emerge from recession with growth turning ositive in the second half of 2009 (nevertheless leaving annual growth for 2009 at -4.1 %). This was exected to be followed by a gradual recovery over the following two years, with EU GDP exanding by 0.7 % in 2010 and by 1.6 % in Desite the resilience shown by labour markets to the recession, emloyment was exected to fall in the EU by 2.3 % in 2009 and by a further 1.2 % in 2010, before a return to marginal growth of 0.3% in This reresented an emloyment contraction of around 7.5 million jobs over the two years , and would have resulted in 4 A negative ga meaning that the target had been exceeded. 5 Note that extraolations to 2010 or 2011 shown in this aer are those made for discussions in Emloyment Committee in the first half of 2010, using the Autumn 2009 economic forecast, and they have not been revised using later forecasts. May 2012 I 9

10 emloyment being down around 7 million in 2011 comared to 2008 levels. At the same time, unemloyment was forecast to rise significantly, otentially reaching 10.3 % by 2010 and remaining around 10.2 % in 2011, u more than 3 s on 2008 levels. Turning to individual Member States, emloyment growth was exected to be negative in most both in 2009 and 2010 (Chart 2). Particularly strong falls in 2009 (of around 6 to 12 %) were foreseen for the Baltic States, Ireland and Sain. Though declines were set to lessen significantly in 2010 in almost all Member States, emloyment was set to contract more strongly in Belgium, Germany, Luxembourg, the Netherlands and Poland. Chart 2: Forecast emloyment growth rates over 2009 and % change EU27 US JP LV IE EE FI LT ES DK NL SI DE SE CZ BE BG LU PL UK FR HU EL AT IT PT CY SK MT RO Source: Commission services (ECFIN), 2009 Autumn economic forecast. The economic recession was therefore exected to largely reverse the rogress achieved in raising emloyment rates since The overall emloyment rate was exected to fall by about 3 ercentage oints between 2008 and 2011, back to around 63% (i.e. almost back to the 2000 level of 62.2%), although the decline in ERs was exected to vary considerably across Member States (Chart 3). Particularly shar declines were forecast in Cyrus (mainly due to demograhic develoments), Ireland, Latvia, Lithuania and Sain. Looking ahead to the economic recovery, because substantial labour hoarding took lace during the recession, articularly in those Member States that relied extensively on short time working arrangements (e.g. Germany) leading to significant declines in labour roductivity, the recovery of roductivity levels is likely to ut a brake on the ace of emloyment growth at least during the initial hase of the recovery. Indeed, most forecasts exect job creation in the early art of the recovery to be subdued, as the widesread resort to reductions in hours worked is first reversed before emloyers look to exand staff levels. May 2012 I 10

11 Chart 3: Forecast change in ERs across EU Member States between 2008 and % of oulation DK NL SE AT UK FI CY DE EE LV SI PT IE CZ EU27 FR LT ES BG LU BE SK GR PL RO IT HU MT Source: DG EMPL calculations based on Nov 2009 EC Economic Forecasts Analysis carried out by DG Economics and Finance (ECFIN) 6 of ast financial and economic crises rovides some tentative indications about the robable longer term imact of the crisis on the level and growth rate of otential outut and the associated imlications for emloyment growth: Given the financial nature of the crisis, the analysis oints to it having both a large negative imact on outut in the short-run and the likelihood of a rolonged eriod of slow growth as economies adjust to a reduced level of otential outut. Past eisodes of financial distress were characterised by sizeable losses in outut - at least twice as large as in more "classical" downturns - and in emloyment. "Classical crises" in general tend to have a ermanent negative effect on the level of GDP, although financial crises can also have a negative imact on long-term growth rates. The emloyment loss - related to the rise in the NAIRU and to reductions in the size of the labour force - is usually not recoued in the decade following the crisis; However, a short recession would not affect the ace of growth of the labour force, leaving otential growth unharmed in the longer run. In contrast, a long and dee recession may cut the otential labour force by discouraging some workers from seeking a job and by reducing migration flows. Moreover, olitical ressure may grow for imlementation of olicies that curtail labour market articiation (e.g. early retirement, curbs on migration flows). The historical background of emloyment rate develoments and the ossible imlications of the recession for future labour market develoments (over the recovery hase and beyond) set the scene for the subsequent analysis of ossible ER targets for Format for the 2020 target Before roosing any ER target for 2020, it was first necessary to carry out an assessment of the continued aroriateness (of the format) of the existing emloyment rate indicator, to see whether it needed adjusting. In articular, several develoments, 6 "Imact of the current economic and financial crisis on otential outut", Euroean Economy Occasional Paers no. 49 (June 2009) May 2012 I 11

12 including EU objectives for furthering educational attainment, articularly for youth, and extending working lives for older workers, have led to concerns over whether the age range (15-64) used for the overall Lisbon emloyment rate target remained aroriate. Given the objectives for raising education levels, lowering school droout rates and taking advantage of higher longevity and imrovements in health conditions to extend labour market articiation among older workers, the merit of adjusting the current age brackets was investigated. 3.1 The imact of rising levels of education During recent decades, rates of formal educational attainment and the qualifications of the workforce have increased across Euroe. Data show that emloyment rates of individuals with higher levels of education are substantially above those of individuals with lower levels (Table 2). Evidence on the close links between education and labour market outcomes has increased olicy awareness and stirred action on these fields - better use of the synergies between education and emloyment is crucial to foster the goals of more and better jobs. Table 2: Trends in emloyment rates by educational attainment ( ) United States Low level (ISCED 0-2) Medium level (ISCED 3-4) High level (ISCED 5-6) EU19 average Low level (ISCED 0-2) Medium level (ISCED 3-4) High level (ISCED 5-6) Source: OECD. Education at a Glance Low level: below uer secondary; Medium level:uer secondary and ost-secondary non-tertiary;high level:tertiary education EU19: AT BE CZ DK FI FR DE GR HU IE IT LU NL PL PT SK ES SE UK With the udating of the Social Agenda 7 and the Communication on New Skills for New Jobs 8, the Commission signalled both the continuity in olicy aims and a new emhasis on a number of olicy areas, inter alia, the ugrading of skills at all levels, romoting emloyability, and matching skills to labour market needs. In May 2009 the Council of Education, Youth and Culture aroved the strategic framework for Euroean cooeration in education and training for the eriod ("ET 2020"). 9 This set a series of Euroean targets to guide the olicy of cooeration in the education and training fields for the eriod to 2020, including that by 2020, the share of year olds with tertiary education should be at least 40%, and that the share of early leavers from education and training should be less than 10%. The simultaneous resence of both ER and education objectives raises the issue of their comatibility, articularly given that in a majority of EU Member States it is rare to accumulate education with work (even art-time). Given the uward trend in levels of educational attainment, ostoning entry in the labour market, together with the thinking exressed in olicy documents favouring furthering education and training objectives, there aeared to be a strong rationale for at least raising the minimum age bracket of the age-range used in the calculation of emloyment rates. 3.2 The age ranges for the ER target Develoments in the age rofile of emloyment rates Over the last decade there has been a shift to older ages in the age rofile of emloyment rates, and this trend is exected to continue over the next decade, esecially for women (Chart 4). While the younger age grou saw emloyment rates decline slightly between 2000 and 2007, rates for all other 5-year grouings 7 COM (2008) 412 final. 8 COM (2008) 868 final. 9 Council Conclusions on a strategic framework for Euroean cooeration in education and training ("ET 2020"), 294 th Education, Youth and Culture Council meeting (Brussels, 12 May 2009). May 2012 I 12

13 increased, most notably for those grous aged 50-54, and Looking forward to 2020, rates for all age grous are exected to rise, but most substantially for the grous aged between 45 and 59. Chart 4: Age rofile of emloyment rates in the EU27 in 2000 and 2007, and 2020 rojections In terms of raising the overall ER, the otential contribution of the age grous and is relatively limited when comared to that of the age grous between 25 and 54, mainly because of the increasing time sent in education activities by the former. Therefore, in order to raise the overall ER, focus should shift to older age grous (55-59 and 60-64) as otentially the ones where rates are substantially below those for the age grou and there is real scoe for raising them. Looking in more detail at age-secific emloyment rates (Chart 5) the distribution of rates by individual year of age highlights certain issues with regard to using the age range as the reference oulation for the emloyment rate target. Of those aged 15 years, only 3.4% were in emloyment in the EU27 in This is lower than the share for all other ages, including all older ages u to and including 74 (3.9%), raising questions on the inclusion of the 15 year grou when others from 65 to 74 are not included yet have higher rates (esecially as with ageing their relative imortance in the oulation will rise in future). For age 16, around one-in-ten is in emloyment, around the same share as those aged 65 or 66 Emloyment rates have risen noticeably between 2000 and 2007 in the older age grou for all ages u to 65. This trend may well continue over the next decade. May 2012 I 13

14 Chart 5: Emloyment rates in the EU27 by individual year of age, 2000 and 2007 % of age oulation ER 2000 ER Source: Eurostat, EU Labour Force Survey The uer age limit Demograhic develoments will continue to have a rofound effect on the EU oulation rofile between 2010 and 2020 (Chart 6), with the eak of the age distribution rising from around years in 2000, to years in 2010 and in In line with this, the age rofile of the oulation aged will also change significantly, with young eole accounting for a lower share (16.3% in 2020 comared to 18.1% in 2010), and older eole aged accounting for a considerably larger share (20.5% in 2020 comared to 18.1% in 2010), meaning that develoments in the older oulation will weigh more in the overall emloyment rate. Furthermore, the number of eole aged will exand substantially in the next decade, u from 46 million in 2010 to almost 55 million by 2020, thereby increasing their relative imortance in the oulation. At the same time, the objective to raise the exit age from the labour force as art of the resonse to demograhic ageing calls into question the use of 64 as the aroriate uer limit for the oulation considered to be of "working age". 10 Figures on the age at which eole exit the labour market and become economically inactive show that, on average, eole within the EU-25 withdrew from the labour force at the age of 60 in Yet due to imrovements in health and living conditions, by the early 2000s life exectancy at the age of 60 had risen on average to around 20 years for men and 24 years for women. This means that currently eole can exect to live substantially beyond the age at which they withdraw from the labour market, which modifies enormously the meaning of retirement. Continuing imrovements in longevity are exected to increase the average duration of retirement even further. 10 The aim to delay the average age of exit from the labour market by 5 years the Stockholm target imlies an average centred around 65 years, which in turn requires that art of the distribution of active older eole be above this age limit. May 2012 I 14

15 Chart 6: Demograhic rofile of EU27 oulation aged u to 79 in 2000, 2010 and Poulation in millions y0 y5 y10 y15 y20 y25 y30 y35 y40 y45 y50 y55 y60 y65 y70 y75 Source: Eurostat, Demograhic data and EUROPOP2008 Year of age Furthermore, in resonse to the need to raise the labour market articiation of older workers, many Member States have carried out reforms of their ension systems (or are in the rocess of doing so), which vary considerably across Member States in terms of the age of eligibility to ension entitlements (see Table I. 1). These reforms have included increases in the age of eligibility for a full ension, bringing retirement ages for women into line with those for men, and tightening the eligibility to early retirement schemes, all of which should have a substantial imact on older workers' emloyment rates over the next decade. In light of the above, the scoe for raising the uer age limit for the ER target was investigated. Data showed that between 2000 and 2007 the emloyment rate for the age grou had risen from 8.8 % to 9.6%. In 2007, the age grou accounted for 24.3 million eole, and emloyment of 2.3 million, while in 2020 it was rojected to account for 29.3 million eole and an overall emloyment level of 2.9 million. This comared with a 2007 emloyment total of 5.7 million for the age grou 15-19, which was rojected to decline to 5.6 million in Although the imortance of the age grou in the working age oulation will therefore increase somewhat, its imact on overall emloyment is foreseen to remain relatively limited. It did not therefore seem relevant to extend the age grou range to cover the whole age grou Nevertheless, consideration was given to ossibly extending the uer age range slightly, by one or two years (e.g. to 65 or 66 years), to better encomass changes in the uer art of the distribution of the working age oulation as older eole's working lives are extended and to reinforce the olicy message of the imortance of active ageing. In the event however, this roosal did not gain the suort of Member States in the Emloyment Committee, who felt it was not the right moment to increase the uer boundary in the definition of working age, and consequently this issue was not exlored further The lower age limit A much stronger case for changing the age boundary alied to the lower age limit for examle a clear argument in favour of raising the lower age bracket is rovided by the very low articiation rates for young cohorts (Table 3). May 2012 I 15

16 Table 3: Particiation and emloyment rates in 2007 in the EU27 Age Brackets PR ER Source: DG EMPL As eole stay longer in education, and as the normal age to which eole stay in fulltime comulsory education may rise in future in some if not all Member States, this calls into question the aroriateness of 15 as the lower bound for the definition of working age. Already, in only 7 EU Member States does full-time comulsory education end by 15 years of age (see Table I. 2). In most (18 Member States), it is already at 16 years of age, while in 2 Member States it is u to 18 years of age. This would suggest a need to raise the lower age limit to at least 16, which would also be more comarable with the lower age boundary used in the US to define the working age oulation. At the same time the share of young eole articiating in education beyond the age of 16 is on the rise and very high for ages below 20 (Chart 7). Indeed, in 2006 around 77% of 18 year-olds were enrolled in education, and just over half of 20 year-olds, while among the age grou just under 86% were in education. With such extremely high levels of articiation in education for those still in their teens, this raises the question of the aroriateness of their inclusion in the bounds of the working age oulation. One consequence of retaining this age grou in the definition for the working age range is that increasing articiation in education (a ositive trend) would automatically lead to falls in the emloyment rate (at least in the short-term). The above suggests that inclusion of the age grou in the definition of the working age oulation, and for ractical uroses even the standard 5-year age band often reorted in statistical data, is becoming increasingly inaroriate. Consequently, consideration was given to raising the lower age limit for the emloyment rate indicator to 20 years, so that the trend in articiation in formal education would not affect emloyment rates so much. For the age grou no longer included in the reference emloyment rate, it was recommended to develo a comlementary indicator for youth neither in emloyment, nor education nor training (NEET). May 2012 I 16

17 Chart 7: Particiation/enrolment in education in the EU27 by age in 2000 and 2007 % of corresonding age oulation Years of age Source: Eurostat, Education statistics The ractical effect of adjusting age bands on emloyment rates and coverage of total emloyment Using annual results from the EU Labour Force Survey, the imact on emloyment rates and coverage of total emloyment of changing the age bands used in determining the reference oulation for emloyment rates was investigated, with the results shown in Table 4. The following oints are highlighted: Adjusting the working age oulation by droing the 15-year age grou and focussing on the range increases the ER by just over 1 ercentage oint with no significant change in the overall coverage of the emloyed oulation. Similarly, raising the lower age bound to 18 raises the ER by 3 ercentage oints, again with no significant change in the overall coverage of the emloyed oulation (still 97.6% of emloyment 15-74). Raising the lower age band to 20 raises the ER by around 5 s (to 70%), and still covers 96% of total emloyment Extending the uer limit to 66 years while retaining the current lower bound (15-66) reduces the ER by 1.5 s, but combined with the droing of the 15 year age grou (16-66) only leads to a 0.5 s reduction, in both cases with a marginal imrovement in overall coverage of emloyment. Extending the uer limit further to 69 has a more marked imact in driving down the ER (by around 4-5 s comared to ending the range at 64 years). In all cases, the level of coverage of total emloyment does not vary substantially. May 2012 I 17

18 Table 4: Effects of using adjusted age bands for the working age oulation, EU27 in 2007 % of total Age band ER emloyment (of age grou 15-74) Source: Eurostat, EU Labour Force Survey, 2007 annual results Based on the receding arguments and the estimates for the imact on emloyment rates and total emloyment coverage of the different age bands considered, the Emloyment Committee indicated a reference for the age band as the reference oulation for the 2020 emloyment rate target. Consequently, Commission Services focused further analysis on that secific age grou. 4. Possible values for the 2020 target Following the direction of the Emloyment Committee, the age range was used to exlore aroriate/feasible values for the emloyment rate target in The analysis took into account not only long-term trends and challenges, but also the exected imact of the recession, which may lead to lasting negative effects on outut and emloyment throughout the next decade. 4.1 Aroach The scoe of the rogress already achieved u to 2008, the scale of the demograhic challenge in coming decades, and the furthering of educational objectives articularly for youth all seemed to call for setting even more ambitious targets for 2020 in order to suort adequate labour inut to the economy. However, while being ambitious it is still necessary to remain credible, esecially in light of recent events. The recession was the most severe since the Great Deression. Significant shortfalls in roduction have emerged and may only be artially recoverable in the coming years. 12 As a result, both the level and growth rate of otential outut may be affected in the coming years, reflecting inter alia a rise in structural unemloyment and/or a reduction in articiation rates. 13 Based on a suly-side methodology develoed by the AWG 14, and using Eurostat's baseline demograhic rojections ublished in 2008 and the Commission's autumn 2009 economic forecast (the latest available at the time of the analysis), an outlook for major 11 However, other age bands (e.g ) were also used to rovide a link to the historical/current situation and carry out sensitivity analysis. 12 Bruegel (2009), 'Euroe's economic riorities Memos to the new Commission' 13 DG ECFIN (2009), 'Imact of the current economic and financial crisis on otential outut', Occasional Paers Jointly roduced by the Ageing Working Grou (AWG) of the Economic Policy Committee and DG ECFIN in the framework of the Ageing Reort. May 2012 I 18

19 labour market variables u to 2020 was generated. Given the uncertainty surrounding the ace of economic recovery and labour market develoments (e.g. the emergence or not of unemloyment hysteresis), three ossible scenarios for labour markets outcomes in the eriod u to 2020 were considered, based on the baseline AWG rojections adjusted to the latest data available at the time of the analysis: i) EU labour markets revert to best historical erformance in terms of unemloyment rates in the 2000s; ii) emergence of hysteresis unemloyment; and iii) unemloyment rates by 2020 decline to minimum (or 'frictional') levels i.e. even below best historical outcomes. 4.2 Eurostat's 2008 baseline scenario for demograhic variables Eurostat's demograhic rojections were released in Aril 2008 (under the reference "EUROPOP2008"). These rojections indicate that in coming decades, the age-structure of the Euroean workforce will undergo dramatic changes due to low fertility, continuous rises in life exectancy and the retirement of the baby-boom generations. Aart from the imact on the labour force, ageing will also have rofound economic, budgetary and social consequences. Changes in the size and age rofile of the oulation deend uon assumtions regarding fertility rates, life exectancy and migration. As far as fertility and mortality rates are concerned, the rojections assume that country-secific rates will converge to that of forerunners over the very long-term i.e. by Furthermore, life exectancy gains are rojected. 16 As regards migration, Eurostat's demograhic rojection assumes a continued but decelerating inward net migration to the EU. 17 According to Eurostat's baseline scenario, in the EU27 the working age oulation (aged 15-64) is exected to decrease by around 1 million eole by 2020 (see Table 5). Table 5 and Chart 8 show that the exected evolution of the working age oulation (WAP) deends on the choice of age brackets, and that the decline in the WAP only begins around the middle of the next decade; first at a slow ace but accelerating after In fact, the moderate decline in the WAP until 2020 is misleading regarding exected develoments after Table 5: Working age oulation in the EU27 by age bands (millions) g g y g ( ) Sources: LFS and Eurostat's 2008 demograhic rojections. 15 Forerunners concerning fertility are: DK (2006:1.83), FI (2006: 1.84), SE (2006: 1.85) as well as UK (2006: 1.84). France, Ireland and Norway are above 1.9. The theoretical convergence of Total Fertility Rate (TFR) was set to 1.85 to be achieved by For mortality rates, the forerunners are France, Italy, Sain and Sweden. 16 In the EU, life exectancy at birth for males is rojected to increase by 8.5 years from 76 in 2008 to 84.5 in For females, life exectancy at birth is rojected to increase by 6.9 years for females, from 82.1 in 2008 to 89 in 2060, imlying a narrowing ga between males and females life exectancy. 17 For the EU as a whole, annual net inflows are rojected to fall from about 1.7 million eole in 2008 (equivalent to 0.33% of the EU oulation) to just under 1 million by 2020 and thereafter to some 0.8 million eole by 2060 (0.16% of the EU oulation). May 2012 I 19

20 Chart 8: WAP in the EU27 (millions) WAP WAP The evolution of the total WAP is, however, an inaroriate indicator of the imact of ageing on socio-economic variables. The structure (or comosition) of the WAP gives a much better measure of the otential imact of ageing (Chart 9). In fact, while the WAP (15-64) declines between 2008 and 2020 by only 0.2% in the EU27, the age distribution of the oulation changes significantly: the ercentage of younger eole (15-29) falls from 28.2% in 2008 to 25.4% in 2020, while that of older eole (50-64) increases from 28.1% to 32.0%. These changes in the structure of the oulation are significant, articularly given the short eriod considered. Chart 9: Age distribution of the working age oulation (15-64) 11.0% 10.0% 9.0% 8.0% % 6.0% 5.0% May 2012 I 20

21 4.3 The methodology for rojecting labour market variables Three labour market rojections for 2020 were calculated based on the 'suly-side' methodology develoed by the AWG 18 for the Ageing Reort. 19 The AWG's labour market rojections follow the age cohort methodology first develoed by the OECD 20, with one modification, which is the use of single year of age instead of five-year age grous. Comared with standard rojections based on the invariance of articiation rates, the cohort-based rojection contains an autonomous increase for the articiation of women referred to as the 'cohort effect' corresonding to the gradual relacement of currently older women, with relatively low articiation rates, by younger women with a much stronger attachment to the labour force 21, but also incororates a negative 'cohort effect' for men because, according to long term trends, their articiation rates have decreased across generations in a large majority of countries, contrary to what is observed for women. Labour market rojections were carried out by country, gender and age grou. The following 13 age grous were considered: 15-19, 20-24, 25-29, 30-34, 35-39, 40-44, 45-49, 50-54, 55-59, 60-64, 65-66, 67-68, and Particiation rates The first ste in calculating emloyment rates is to roject articiation rates (Chart 10). Chart 10: Labour market rojections (the main stes in colour/darker) EU LFS Cohort effects and Pension reforms Particiation rates Demograhic scenario Labour force Unemloyment rate assumtions Unemloyment Emloyment Emloyment rates Secifically, the methodology to roject articiation rates makes the following major assumtions: 18 The Ageing Working Grou was established by the Economic and Policy Committee. 19 The 2009 Ageing Reort: Underlying assumtions and rojection methodologies for the EU-27 Member States ( ), Euroean Economy 7/ Burniaux, Duval, and Jaumotte (2003), 'Coing with ageing: a dynamic aroach to quantify the imact of alternative olicy otions on future labour suly in OECD countries', OECD Economic Deartment WP n The model used in the AWG's rojection is based on the assumtion that lifetime articiation rofiles in the future arallel those observed in the ast. This imlies the assumtion that the entry and exit rates calculated for the latest available cohorts ( ) are ket constant in the future. Comared with a static baseline, this method imlies a gradual increase of future articiation rates for women, articularly for those aged 35 and over. May 2012 I 21

22 Particiation rates are calculated by single age and sex, by using the entry/exit rate calculated on the basis of the average of articiation rates observed over the eriod Therefore, articiation rates should be interreted as averages over a 'normal' economic cycle. The exit rates of older workers (aged 55-71) were adjusted to take into account recent and lagged effects of ension reforms in 20 EU Member States. In fact, a very imortant feature of the AWG's labour market rojection is that it takes into consideration the otential effects of recently enacted ension reforms in EU Member States, including measures to be hased in only gradually, on the articiation of older workers (aged 55-71). Table 6 resents the resulting articiation rates. Values for 2008 are historical EU LFS data. Values for 2011 reflect the Euroean Commission's autumn 2009 forecast (using 2008 as baseline). Values for 2020 are taken from the AWG rojection. These indicate that in the EU27 articiation rates are exected to increase by 2.6 ercentage oints () between 2011 and 2020 (age band 15-64), reflecting increases of 4 and 10, resectively, in the articiation of women and older workers (aged 55-64). The latter reflects both cohort effects and the imact of ension reforms. May 2012 I 22

23 Table 6: Particiation rates Age band Age band Age band COUNTRY YEAR Men Women Total Men Women Total Men Women Total AT BE BG CY CZ DE DK EE ES FI FR GR HU IE IT LT LU LV MT NL PL PT RO SE SI SK UK EU Sources: EU LFS (2008), DG Ecfin autumn 2009 forecast (2011), and AWG rojections (2020). May 2012 I 23