Youth & The UK Labour Market. March 15th. Jonathan Wadsworth. Royal Holloway College, CEP LSE, CREAM UCL, MAC and IZA Bonn

Transcription

1 1 Youth & The UK Labour Market March 15th Jonathan Wadsworth Royal Holloway College, CEP LSE, CREAM UCL, MAC and IZA Bonn

2 2 Labour Market Performance of Young Adults Young people are typically always at a disadvantage in the labour market. Historically youth unemployment rates is twice that of the adult rate Didn t used to worry unduly about youth because while more likely to experience unemployment, typically duration of any jobless spell was much shorter (more likely to be hired than older workers) The slump - which inflicted a larger cumulative loss of output than any other post-war downturn - has led to concerns that youth unemployment has risen to such an extent that there is a risk of a lost generation We are right to worry about scarring effects for young adults, but is the youth labour market really different (worse) this time round?

3 3 Long known that combination of characteristics matter for chances of being jobless A graph of the relative chances (% point gap in unemployment-pop ratio) of being unemployed for certain key characteristics shows young adults worse off unemployment pop.ratio differentials by type % point gap female north yorks e mids e angl London s west w mids n west Wales Scotland age graduate intermediate low immigrant eth. min.

4 But this is from

5 5 From 2012 and 1993 again unemployment pop. ratio differentials by type unemployment pop.ratio differentials by type % point gap % point gap female north yorks e mids e angl London s west w mids n west Wales Scotland age graduate intermediate low immigrant eth. min. female north yorks e mids e angl London s west w mids n west Wales Scotland age graduate intermediate low immigrant eth. min. So many of the same factors that were relevant for labour market performance in the 1990s (and earlier) are still relevant now and the relative importance of age and also other characteristics (education, ethnicity gender) hasn t changed much either. Despite the causes of the downturn being different, the manifestation of its effects is much the same

6 So is the current concern not justified? 6

7 Youth Labour Market (16-24) year unemp. rate w. students unemp:pop. ratio NEET Youth unemployment rate is now higher than in previous 2 recessions but both u. pop rate and NEET rate still below 1990s peak (which in turn lower than 1980s)

8 8 Students complicate the youth labour market statistics 7.3 million young adults million in full-time education share in age group Share of students in Youth Population year 1) Some of them (around 300,000) appear legitimately - in the unemployment count at the same time as being recorded as in full-time education 2) Large (partly counter cyclical) rise in student numbers also shrinks the youth labour force and exaggerates the youth unemployment rate u =U/(E+U)

9 9 thousands e e+06 Students in the Unemployment Count year rate Students in the Unemployment Rate year with students no students with students no students Around 300,000 students now in the unemployment count (1/3 of total) Adds about 2 percentage points to the youth unemployment rate

10 10 The Effects of Students & Labour Force Participation on the Unemployment Rate Unemployment Participation rate, Unemp. rate, U/(U+E) (1) (U+E)/(U+E+N) (2) Population ratio U/(U+E+N) (1)*(2) UK Spain France Italy Germany US Source: OECD for 2011 age Author s calculations UK has a very high youth labour market participation rate So the poor performance of youth when measured by the unemployment rate in France, Spain (and the UK) is exaggerated by low levels of labour market participation which is in turn in the main caused by higher levels of participation in tertiary education

11 11 Feeds into longstanding issue regarding ethnicity and youth unemployment Unemployment Rate 2012 q3: Men White Mixed/Multiple ethnic groups Bangladeshi Any other Asian background Pakistani Chinese Indian Other ethnic group Black/African/Caribbean/Black British unemployment rate

12 12 But part of the reason is that labour force participation is lower among many ethnic minorities because participation in education is higher Students in Youth Pop q3: Men White Mixed/Multiple ethnic groups Other ethnic group Black/African/Caribbean/Black British Indian Bangladeshi Pakistani Any other Asian background Chinese Students in (UK-Born) Youth Pop q3: Men White Black/African/Caribbean/Black British Mixed/Multiple ethnic groups Bangladeshi Indian Other ethnic group Chinese Pakistani Any other Asian background student share student share And a smaller labour force exaggerates the unemployment rate

13 13 Probably better to look at NEET rate when looking at youth labour market NEET Rate 2012 q3: Men Chinese Any other Asian background Bangladeshi Mixed/Multiple ethnic groups White Indian Pakistani Other ethnic group Black/African/Caribbean/Black British NEET Rate (UK-Born) 2012 q3: Men Any other Asian background Bangladeshi Pakistani Mixed/Multiple ethnic groups White Indian Chinese Other ethnic group Black/African/Caribbean/Black British neet share neet share So not nearly as much difference as when look at unemployment rate

14 14 And none when account for sampling variation (standard sampling error is around +- 6 percentage points for minority groups) Sampling errors do include a big range of values so need to be careful making comparisons across small samples NEET Rate ETHUKEUL== White ETHUKEUL== Mixed/Multiple ethnic groups ETHUKEUL== Indian ETHUKEUL== Pakistani ETHUKEUL== Bangladeshi ETHUKEUL== Chinese ETHUKEUL== Any other Asian background ETHUKEUL== Black/African/Caribbean/Black British ETHUKEUL== Other ethnic group ethnicity Male Female On this basis only young Bangladeshi women statistically significantly different from young white women

15 Has the performance of youth deteriorated relative to other age groups? 15

16 16 Relative to age u. pop. relative ratio Total with students ex. students 50+ Relative with students u:pop ratio is higher (though peaked in 2007) Relative without students u:pop ratio is still not as bad as 1980s (worse than 1990s) (Over 50s doing better)

17 17 So hard to argue recent shift in relative labour market position of youth on this basis but.

18 18 Neet rate has worsened relative to other age groups (mainly because of continued rise in participation among women 25+) neet rate relative to neet rate year

19 19 So only in relative but not in absolute terms could argue that youth labour market is unprecedented (and now NEET rate is equal to and below 50+) ¾ of all NEETS have GCSE or less 50% of all with GCSE or less are NEET (2012)

20 20 History shows that youth unemployment is dominated by large inflows and large outflows (so might experience lots unemployment but, typically, don t stay unemployed very long and so, arguably, less of a concern) Any different this time?

21 21 Current spells of youth unemployment dominated by short-term spells, (much more so than for other ages) Distribution of Unemployment by Duration: All q3 Fraction Distribution of Unemployment by Duration: Others 2012 q3 Fraction <3m 3-6m 6-12m 12-18m 18-24m 2-3yr 3-4 yr 4-5yr 5yr + Duration of Ongoing Spell <3m 3-6m 6-12m 12-18m 18-24m 2-3yr 3-4 yr 4-5yr 5yr + Duration of Ongoing Spell (q3 boosted by school/college leavers)

22 22 but most of total unemployment days accounted for by the minority who are very long-term unemployed If add total days lost to youth unemployment then the currently 1 million unemployed have been unemployed for around 9 million months (270 million days) of which <15% is accounted for by the 55% who are short-term unemployed (< 6 months) Distribution of Total Days Lost to Youth Unemployment Distribution of Total Days Lost to Unemployment Share of Total Youth Unemployed Months <3m 3-6m 6-12m 12-18m 18-24m 2-3yr 3-4 yr 4-5yr 5yr + Share of Total Unemployed Months <3m 3-6m 6-12m 12-18m 18-24m 2-3yr 3-4 yr 4-5yr 5yr +

23 23 and around half of young neets have never worked Distribution of Non-Employment by Duration: Young Adults Fraction <3m 3-6m 6-12m 1-2 yr 2-3 yr 3-4 yr 4-5 yr 5yr + never worked Duration of Ongoing Spell

24 24 So need to worry about a) Getting young people work experience and/or jobs b) build-up of long-term unemployment (among youth) In and of itself and also because of link with scarring

25 25 Effect of youth unemployment on later experience of unemployment (BCS) Group Type at Age 23 % of sample No spell of unemployment Average percentage time spent unemployed age (% of group with any unemployment in this interval) (7.5) Average percentage time spent inactive age (% of group with any inactivity in this interval) 2.3 (9.6) 1-5 months of unemployment 6-12 months of unemployment (13.8) (21.4) 3.7 (15.6) 7.1 (24.6) 13+ months of unemployment (40.0) 22.9 (46.8) Source: Gregg (2007) Scarring effects bad but (historically) confined to a minority of youth unemployed - no evidence yet exists for this time round

26 26 Long-term unemployment rates also narrowing by age groups Though relative to previous recessions youth share still not quite as bad Share of Unemployed Age Group with Spell 12 months + LTU (12 mth +) share year

27 27 However rising share of young neets have never worked share of neet never worked year (45% of Neets means around 10% of all never worked)

28 28 Are Youth being substituted for Older Workers? - probably not (since they, largely, do different jobs) Occupation Occupation 2012 Age 50+ Managers, Directors and Senior Officials Professional Occupations Associate Professional and Technical Occupations Administrative and Secretarial Occupations Skilled Trades Occupations Caring, Leisure and Other Service Occupations Sales and Customer Service Occupations Process, Plant and Machine Operatives Elementary Occupations Managers, Directors and Senior Officials Professional Occupations Associate Professional and Technical Occupations Administrative and Secretarial Occupations Skilled Trades Occupations Caring, Leisure and Other Service Occupations Sales and Customer Service Occupations Process, Plant and Machine Operatives Elementary Occupations group size e e+06 group size Estimated elasticity of substitution between age groups = 7 (age groups not perfect substitutes)

29 29 Less skilled neet rate not very well correlated with local area performance of over 50s Low Skilled Youth UK-Born Change in NEET Rate Change in Age 50+ Emp. rate actual value fitted line

30 By immigrants? 30

31 31 Probably not Local Area Change in less skilled youth unemployment v. Change in Immigration Share Low Skilled UK-Born Youth Change in Unemp. Rate Change in Immigrant Share actual value fitted line ΔUmp. Rate Youth = ΔImmig. Share (if anything negative) (0.01) (0.19)

32 32 Nor does s it matter if use EU-migrants (pret-effect) Low Skilled UK-Born Youth Change in Unemp. Rate Change in EU Immigrant Share actual value fitted line (or use the employment rate)

33 So is it the minimum wage? 33

34 34 Probably not Low Skilled Youth UK-Born Low Skilled Youth UK-Born Change in NEET Rate Change in NEET Rate Change in Adult Kaitz (Min. Wage Bite) Change in Youth Kaitz (Youth Min. Wage Bite) actual value fitted line actual value fitted line Little correlation between changes in lees skilled youth neet and change in bite of either adult or youth minimum (or over the period or if use unemp rate )

35 35 Cohort size? count of sex e and over Probably not (larger youth cohorts in the 1980s)

36 36 What it is correlated with is the overall local area performance (neet rates are much lower in areas that are doing ok) Low Skilled Youth UK-Born Change in NEET Rate Change in local area Emp. rate (ex neet) actual value fitted line

37 Going forward 37

38 38 Just as with unemployment chances of being neet vary with combinations of region, age, gender and above all education Youth neet rate differentials by type 2012 % point gap female north yorks e mids e angl London s west w mids n west Wales Scotland age age graduate intermediate low immigrant eth. min.

39 History suggests that won t get anywhere without growth in excess of 2% a year 39

40 40 Okun s Law suggests 2.3% growth needed before aggregate unemployment rate falls quarterly change in unemp. rate annualised % change in GDP Fitted Line Change in Unemployment Rate = *%Change in Real GDP

41 41 but Okun s Law suggests 3% growth may be needed before youth unemployment rate falls quarterly change in youth unemp. rate annualised % change in GDP Fitted Line Change in Unemployment Rate = *%Change in Real GDP (the more balanced the growth the better the likely outcome) Source: ONS quarterly data, author s calculation

42 42 Young adults are more skilled in absolute terms - than in the past - this should help % in full-time education % with GCSE or lower (36% 16-19) 75 (66% 20-24) (71% 16-19) 42 (27% 20-24) But main problems are for those with lower qualifications

43 43 Where will jobs come from? -traditional employers of less skilled youth labour are Distribution of Employment by Industry: Less Skilled Youth 2012 Distribution of Employment by Industry: Skilled Youth 2012 Fraction Fraction Agric Energy Manf Const Retail Transr Finan Public Oth.Ser Sector Agric Energy Manf Const Retail Transr Finan Public Oth.Ser Sector

44 44 And older adults Distribution of Employment by Industry: Less Skilled Age Fraction Fraction Distribution of Employment by Industry: Skilled Age Agric Energy Manf Const Retail Transr Finan Public Oth.Ser Sector Agric Energy Manf Const Retail Transr Finan Public Oth.Ser Sector

45 45 But recoveries do help (though probably not enough if not balanced) Change in relative employment rate performance of marginal groups across area labour markets Total Change in area employment Men 50+ Low Quals. Female Lone parents low quals. Male low quals.* Best Worst Best Worst Best Worst Best Worst (5 p p+) (< 0 p p) * neet rate So individuals with similar characteristics get a differential benefit depending on aggregate local area performance (best = tyne & wear, south yorks, strathclyde) (worst=west yorks, west mids met., west mids)

46 46 Conclusions Not much evidence to suggest that current problems of youth labour market bad though they are are the result of anything other than the severe downturn (and we have been in similar if not worse position regarding youth in past 2 downturns) So if things start to get better then prospects for youth labour market should improve (at least for majority) but probably not until growth in excess of 2% So may be a case of waiting/inducing demand and in the meantime ensuring schemes keep young people in touch with the labour market (with particular focus on less skilled since will benefit last from any upturn and here targeted intervention is probably needed) -so keep the schemes (maintain outflows from U and N) but don t forget about Substitution (long-term unemployed just sub. for short-term) Displacement (of firms not benefitting from scheme) Deadweight loss (pay for something happened anyway) (expensive in a downturn lots of clients )