www.skope.ox.ac.uk ‘Have earnings polarised in the UK? Craig Holmes Pembroke College, Oxford University and SKOPE ISER, University of Essex, October 21 st 2013
Feb 24, 2016
www.skope.ox.ac.uk
‘Have earnings polarised in the UK?
Craig HolmesPembroke College, Oxford University and SKOPE
ISER, University of Essex, October 21st 2013
www.skope.ox.ac.uk
Outline
• Background on the hourglass labour market• Polarisation and earnings distributions – some theory• Methodology – earnings distribution decomposition• Data• Decomposition results• Discussion and future work
www.skope.ox.ac.uk
Background
• Routinisation hypothesis (Autor, Levy and Murnane, 2003):– Refinement of SBTC - technology related to tasks, not skills– Routine tasks substitutable for computer capital– Growth in non-routine jobs, decline in routine jobs
• Polarization hypothesis (Goos and Manning, 2007)– Routine occupations found in middle of income distribution– Non-routine occupations found at top and bottom of distribution– Ranking of jobs based on initial wages
www.skope.ox.ac.uk
Background
• Goos and Manning (2007) – 1979-1999:
www.skope.ox.ac.uk
Background
• Similar results observed in:– US (Autor, Katz and Kearney, 2006; Caranci and Jones, 2011) – Germany (Spitz-Oener, 2006; Oesch and Rodríguez Menés, 2011)– Spain and Switzerland (Oesch and Rodríguez Menés, 2011) – Across Europe (Goos, Manning and Salomons, 2009)
• Other explanations have been put forward:– Offshoring– Growing wage inequality and demand for services
www.skope.ox.ac.uk
Background• Wage inequality in the UK has risen since the 1980s
www.skope.ox.ac.uk
Background• Earnings growth by percentile, UK
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 10.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
35.0%
40.0%
1987-2001
1994-2007
Percentile
Real
ear
ning
s gro
wth
www.skope.ox.ac.uk
Background
• More high-wage and low-wage jobs More inequality
Year Jobs earning below 2/3 * median hourly
wage
Jobs earning above 1.5* median hourly
wageInitial (1987) 20.2% 23.4%Final (2001) 23.0% 25.6%Initial (1994) 22.6% 25.2%Final (2007) 21.3% 25.9%
www.skope.ox.ac.uk
Background
• Two main research questions:– To what extent has the shift towards non-routine employment
decreased the number of middle wage jobs / increased wage inequality
– Why, given that, has earnings distribution polarisation halted since mid 1990s?
www.skope.ox.ac.uk
Polarisation and earnings distributions
• The shift away from routine work should increase the number of high-wage and low-wage jobs, everything else being equal
0 2 4 6 8 10 12 140.00
0.05
0.10
0.15
0.20
0.25
0.30
All
Routine
High skill non routine
Low skill non routine
Gross hourly wage, 1987
www.skope.ox.ac.uk
Polarisation and earnings distributions
• However, wage structure of occupations unlikely to remain constant
• Autor, Katz and Kearney (2006) – relative wage of routine occupations falls– “Wage polarisation” – a US phenomenon?
• Wage differences between different non-routine occupations (Williams, 2012)– Other compositional changes – more educated workforce, lower union
membership, greater female participation– Non-uniform increase in demand for non-routine tasks?– Change in returns to other characteristics
www.skope.ox.ac.uk
Methodology• Standard quantile regressions compute quantiles of a
distribution conditional on explanatory variables• However, we need to decompose unconditional distributions• Firpo, Fortin and Lemieux (2009) – two stage approach
1. Estimate a counterfactual distribution via reweighting initial distribution
2. Use re-centered influence functions to estimate distributional statistics (such as percentiles) as a linear expression of main explanatory variables
www.skope.ox.ac.uk
Methodology• Data:
– N observations, N0 from initial distribution, N1 from final distribution– Ti = 1 if from final distribution, i = 1,...,N. Pr(Ti) = p
• Data can be reweighted• Reweighting:
– where p(X) = Pr (T=1|X)
)(1)(1
)1&Pr()1|Pr( 00
XpXp
pT
EF
pTyYTyYF
yYC
C
www.skope.ox.ac.uk
Methodology• This counterfactual can be used to decompose wage and
composition effects of a distributional statistic:
• An recentered influence function of v(F) measures its sensitivity to each observation, where E(RIF) = v(F)– Assume a linear projection of RIF onto X:
– where j = {0, C, 1}
CW
CC
vvFvFvFvFvv
01
vjj
vj
vj XRIF
www.skope.ox.ac.uk
A quantile regression approach• Hence:
• This is a more general case of the Blinder-Oaxaca decomposition, where v(F) is the mean.
CW TXEv 11|
00|1| TXETXEvC
www.skope.ox.ac.uk
Data• Family Expenditure Survey, 1987-2001
– Around 10,000 observations each year– Usual gross pay and usual hours of work– Education – year left FT education four levels– Union membership – subscription fees>0
• Quarterly Labour Force Survey, 1994-2007– Around 150,000 observations each quarter (5 quarter membership)– Gross hourly pay directly reported– Educational qualifications– Union membership directly reported
www.skope.ox.ac.uk
Data 1987 2001 1994 2007
Female 47.3% 50.3% 50.7% 52.1%Union membership 29.0% 15.3% 36.9% 31.5%Works part-time 23.6% 23.3% 26.3% 26.6%University graduates 9.4% 16.9% 13.4% 23.4%No qualifications 36.1% 18.6% 17.2% 8.3%Experience < 5 years 11.9% 9.4% 7.7% 8.2%Experience > 20 years 49.1% 53.5% 57.8% 53.8% Professional 11.2% 12.8% 11.9% 14.4%
Managerial 7.3% 11.7% 12.5% 14.9%
Intermediate 10.1% 13.7% 14.0% 15.8%
Manual Routine 36.1% 26.4% 26.9% 19.9%12.2%Admin Routine 19.7% 15.2% 14.7%
Manual Non-routine 1.8% 0.8% 0.9% 1.1%
Service 13.8% 19.4% 17.9% 21.7%
N 7253 5908 32355 54098
www.skope.ox.ac.uk
Composition and wage effects• FES, 1987-2001:
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
-5.0%
0.0%
5.0%
10.0%
15.0%
20.0%
Composition
Wage structure
Percentile
Real
wag
e gr
owth
, 198
7-20
01
www.skope.ox.ac.uk
Composition and wage effects• LFS 1994-2007:
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.000.0%
5.0%
10.0%
15.0%
20.0%
25.0%
Composition
Wage structure
Percentile
Real
wag
e gr
owth
, 199
4-20
07
www.skope.ox.ac.uk
Composition and wage effects• Both periods find compositional changes decreasing the
number of middle-wage jobs• Wage structure changes reverse this – partially in the 1987
and 2001, and completely between 1994 and 2007Year Jobs earning below
2/3 * median hourly wage
Jobs earning above 1.5* median hourly
wageInitial (1987) 20.2% 23.4%Composition effects only 24.0% 27.1%Final (2001) 23.0% 25.6%Initial (1994) 22.6% 25.2%Composition effects only 25.2% 27.3%Final (2007) 21.3% 25.9%
www.skope.ox.ac.uk
Individual composition effects
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
-5.0%
0.0%
5.0%
10.0%
15.0%
20.0%
Total estimated composition
Occupations
Education
Unions
Female
Percentile
Real
wag
e gr
owth
, 198
7-20
01
www.skope.ox.ac.uk
Individual composition effects
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
-2.0%
0.0%
2.0%
4.0%
6.0%
8.0%
10.0%
12.0%
14.0%
Total estimated composition
Occupation
Education
Unions
Female
Percentile
Real
wag
e gr
owth
, 199
4-20
07
www.skope.ox.ac.uk
The wage structure - aggregate
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
-15.0%
-10.0%
-5.0%
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
Wage structure Occupation Education
Unions Female Constant
Percentile
Real
wag
e gr
owth
, 199
4-20
07
www.skope.ox.ac.uk
The wage structure - occupations
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
-2.00%
-1.00%
0.00%
1.00%
2.00%
3.00%
4.00%
5.00%
6.00%Professional
Managerial
Intermediate
Routine manual
Service
Percentile
Real
wag
e gr
owth
, 199
4-20
07
www.skope.ox.ac.uk
The wage structure - education
0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
-3.0%
-2.0%
-1.0%
0.0%
1.0%
2.0%
3.0%
Degree
Post compulsory
Percentile
Real
wag
e gr
owth
, 199
4-20
07
www.skope.ox.ac.uk
Discussion• Wage structure reduces increase in inequality, despite change
in composition on the workforce• However, hard to interpret as educational or occupational
opportunities pulling the middle up – despite the “room at the top” mindset of policymakers
• An alternative interpretation – downward sloping wage structure is a ‘correction’ of compositional changes – not as many people in high wage jobs as we’d predict
www.skope.ox.ac.uk
Discussion• Minimum wage has clearly having an effect at low-end• However, other wage structure effects also helping to reduce
lower-tail inequality e.g. male-female wage gaps, relative pay of service jobs
www.skope.ox.ac.uk
Discussion• Increasingly heterogeneous occupational groups
Low pay Middle pay High pay
-5.00%
-4.00%
-3.00%
-2.00%
-1.00%
0.00%
1.00%
2.00%
3.00%
Professionals Managerial
Intermediate Manual routine
Admin routine Service
Chan
ge in
em
ploy
men
t sha
re, 1
994-
2007
www.skope.ox.ac.uk
Discussion• Unrelated to educational attainment? Graduates only:
Low pay Middle pay High pay
-5.00%
-4.00%
-3.00%
-2.00%
-1.00%
0.00%
1.00%
2.00%
Professionals Managerial
Intermediate Manual routine
Admin routine Service
Chan
ge in
em
ploy
men
t sha
re, 1
994-
2007
www.skope.ox.ac.uk
Discussion• This could reflect a supply problem if it reflects quality of
graduates• Could also reflect suitability of university route into labour
market vs. vocational education• Can not ignore changes on the demand side – in particular,
are technology and skilled labour always complements?• Brown, Lauder and Ashton (2011):
– “Knowledge work” “Working knowledge”– “Digital Taylorism” – deskilling of high skill work– “War for Talent” – high premium paid for small pool of graduates at
top universities
www.skope.ox.ac.uk
Discussion• So far, defined high wage job as a fixed multiple of median pay• However, the size of the wage spectrum is relevant• Very high pay is increasing, even while high pay remains
constant• Some higher paid workers move closer in relative terms to the
middle as the top experience very rapid wage growth• What is an appropriate cut-off for these groups?
www.skope.ox.ac.uk
Conclusion• Main points:
– Earnings distributions polarised during 1980s and 1990s– This polarisation was less than compositional shifts would have
predicted– Middle paying jobs stop declining in middle of 1990s, despite
continuation of compositional shifts• Areas for future investigation:
– Educational attainment and occupational demands do not seem to be offering opportunities to narrow earnings inequality
– What happens to occupational mobility, particularly at low end?– What has happened since 2007?
www.skope.ox.ac.uk
Contact DetailsCraig Holmes
Pembroke College, Oxford, andESRC Centre on Skills, Knowledge and Organisational
Performance (SKOPE),
Email: [email protected]