Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
© Institute for Fiscal Studies
Income Shocks, Assets and Labour Supply: Understanding Inequality Dynamics
Richard Blundell University College London and Institute for Fiscal Studies
Thanks to Jonathan Cribb, Ben Etheridge, Andy Hood, Wenchao Jin, Robert
Joyce, Cormac O’Dea, Luigi Pistaferri, Ian Preston, and Itay Suporta
This research is funded by the ERSC Centre (CPP) at IFS.
MISOC Workshop September 15th 2014
Motivation • Many of the key determinants in the trends in income inequality and in
overall living standards over the past 25 years, including since the financial crisis, have been driven by changes in the labour market, including, for the UK: – huge increase in entry cohorts with at least a BA degree during the 1990s
and early 2000s, – large relaHve rise in top ‘earnings’ percenHle since the early 1990s, – dramaHc fall in real wages since 2008, ….
• To which we can add changes in savings and asset prices, in parHcular housing; and reforms to taxes and welfare benefits.
• Aim: To layout a framework for examining the role of interacHons between labour market shocks, asset markets shocks and changes to the tax and welfare system, in the evoluHon of inequality, – allowing an interpretaHon of what has happened in the UK, and also in the US and elsewhere.
Inequality over the Working Life • Begin by focussing on the labour market as a primary source of shocks
and examine the links between the various definiHons of economic inequality:
• wages => earnings => joint earnings => income => consumpHon – the link between the various measures of inequality is mediated by
mulHple `insurance' mechanisms including:
1. (Family) labour supply (wages → earnings → joint earnings) 2. Taxes and welfare benefits: (earnings → income) 3. Assets: Saving and borrowing (income → consumpHon) 4. Informal contracts, gi^s, etc. • But how important are each of these mechanisms? • How do they change over the life-‐cycle and the business cycle? • How should we design policies to best insure these shocks? First look at income and consumpHon inequality by year and by cohort….
Income DistribuHon by Year -‐ UK
Source: Blundell and O’Dea (2014).
0
200
400
600
800
1000
1200
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
mean
p90
p75
p50
p25
p10
Note: something is happening around 2000, then 2008!
ConsumpHon DistribuHon by Year -‐ UK
0
100
200
300
400
500
600
700
800
900 1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
mean
p90
p75
p50
p25
p10
Source: Blundell and O’Dea (2014).
Note: ConsumpHon started falling back earlier than income in the recent recession….
Income Inequality by Year and Birth Cohort -‐ UK
Source: Blundell and O’Dea, 2014
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
Variance, Log Income
1930s
1940s
1950s
1960s
1970s
1980s
Notice something is happening just after 2000; … much more noticeable in the US.
ConsumpHon Inequality by Year and Birth Cohort -‐ UK
Source: Blundell and O’Dea (2014).
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
Variance, Log Consump4on
1930s
1940s
1950s
1960s
1970s
1980s
ConsumpHon Inequality by Age and Birth Cohort -‐ UK
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
0.40
0.45
21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 Age
1930s
1940s
1950s
1960s
1970s
1980s
Notes: younger cohorts entering with much larger inequality than previous cohorts….
Source: Blundell and O’Dea (2014).
Income Inequality by Age and Birth Cohort -‐ UK
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 Age
1930s
1940s
1950s
1960s
1970s
1980s
Source: Blundell and O’Dea (2014).
Source: Attanasio and Pistaferri (2014)
.1.1
5.2
.25
.3.3
5.4
var(l
ntot
cons
)
20 25 30 35 40 45 50 55 60 65 70age
Born 1930s Born 1940sBorn 1950s Born 1960s
ConsumpHon Inequality by Year and Birth Cohort -‐ US
Inequality dynamics • Similar episodes of growth in income inequality can have very
different dynamic underpinnings, with different policy implicaHons. • Studying the the relaHonship between inequality in income,
consumpHon, wages and earnings can bring some structure to the key features of inequality.
• A ‘shock’ to the earnings of one family member may: – generate reacHons in the labour supply of other family members, – be compensated, in part, through the welfare benefit system, – perhaps also through drawing on savings and net equity, – and even through informal networks.
• This is the moHvaHng idea for the Blundell, Pistaferri and Suporta (2013) study, which uses data from the new PSID, 1996 – 2010, – new panel data measures of consumpHon, earnings and assets; – and previous studies using mixtures of panel and c-‐s data.
Consumption inequality dynamics • BP (QJE 1998) and BPP(AER 2008) approach relates transitory and persistent
log income shocks to log consumpHon. In a simplified form:
Δ lnYit ≈ ν it +Δξitpermanent income shocks
transitory income shocks
Impact of shocks to asset prices, etc.
Ratio of net assets to net assets plus human capital wealth – insurance value of net assets
Δ lnCit ≈ Γit +φ(1−π it−1)ν it +ψγ tL (1−π it−1)ξit +εit
• Relates the evoluHon of the income distribuHon to consumpHon distribuHon – Persistent shocks will, in general, have a larger impact on consumpHon; – ϕ and ψ allow for excess smoothness and excess sensiHvity; and for net savings to be
used to ‘insure’ shocks.
• Income inequality episodes differ in underlying persistence/durability – Using variance of log for UK and US find strong growth in variance of persistent shocks
in early/mid 1980s for all cohorts and early/mid 1990s for certain key working age cohorts.
Born 1940s
0
0.1
0.2
0.3
0.4
0.5
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
Var(c) Var(y) Cov(c,y)
Born 1950s
0
0.1
0.2
0.3
0.4
0.5
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
Var(c) Var(y) Cov(c,y)
Source: Blundell and Preston (1998)
For example, birth cohort inequality:
Implied variance of permanent shocks by cohort – Britain The ‘spike’ in persistent shocks - very similar in US, BPP (AER, 2008)
Source: Blundell and Preston (1998)
Consumption inequality dynamics with family earnings • Blundell, Pistaferri and Suporta (2013) generalise this framework to allow
for mulHple earnings in a family – Wage/earnings shocks can be correlated across family members – Labour supply can respond across family members – Taxes and welfare benefits deliver net current period income – Labour supplies and consumpHon can be non-‐separable
• Some findings: – a key role for family labour supply in providing insurance for adverse shocks,
especially for those with low/no net assets,
– substanHal role for taxes and welfare benefits, and – once family labour supply, taxes, welfare benefits and savings are taken account,
there is lirle role for any other insurance mechanisms.
• Recent recession – assets and real wages suffer an adverse shock: – consumpHon falls for non-‐durables/necessiHes, largely across the distribuHon,
implying strong persistent/uninsurable shocks;
– labour supply of women and older workers is maintained, even enhanced; – ‘double whammy’ implies strong role for welfare benefits and tax credits.
For the rest of the talk….
Focus mostly on the UK:
1. Earnings: Wage Changes, Employment (and ProducHvity) 2. Incomes: Working-‐age (and Pensioners) 3. ConsumpHon: Durable and Non-‐durable Expenditures – think about what they tell us about the process of inequality and living
standards since the recession in the context of the ‘insurance’ model,
– briefly examine the prospects for the future. -‐ the key idea is to use the link between earnings, incomes and consumpHon to build a more detailed picture of the process and prospects for inequality.
© InsHtute for Fiscal Studies
1. Earnings: Wages and Employment
• Average real hourly wages fallen since this recession began. – even though workforce composiHon has shi^ed towards more
producHve types,
• Real wage falls have occurred within individuals: – unprecedentedly high proporHons of employees experienced nominal
wage freezes,
• The educaHon premium survived the large increase in those with BAs – but real wages have fallen for all groups since the recession.
• ‘EffecHve’ and actual labour supply is higher now than during previous recessions
– due partly welfare policy changes and partly to wealth/real wage shocks.
© InsHtute for Fiscal Studies
Real Median Wage by EducaHon, Men
Source: Blundell, Green and Jin (2014)
BA or above proporHon by cohort, all
Source: Blundell, Green and Jin (2014)
ProporHon with degrees or above by age: all
Source: Blundell, Green and Jin (2014)
ProporHon with degrees or above by age: women
Source: Blundell, Green and Jin (2014)
RaHo of BA (S3) median wage to that of S2
Source: Blundell, Green and Jin (2014) Remarkably… no cohort effects! BA premium stayed constant, even through the recession.
Average real hourly wage by qualificaHon among young people
0.00
2.00
4.00
6.00
8.00
10.00
12.00
14.00
18-‐21, GCSEs or above
18-‐21, no GCSEs or equilvants
22-‐24, degree or above
22-‐24, GCSEs, no degrees
22-‐24, no GCSEs
Source: Blundell, Crawford and Jin (2014)
% change to real hourly wage since last year, by period
Source: Blundell, Crawford and Jin (2014)
Employment and labour market parHcipaHon
• Some increase of parHcipaHon can be arributed policy changes, e.g. : – Labour supply has increased among lone parents, partly as a
result of job search condiHons arached to benefit claims
– Older workers are reHring later, partly as a result of increased SPA for women
Change to lone mothers’ parHcipaHon rate since policy change
-‐8% -‐6% -‐4% -‐2% 0% 2% 4% 6% 8% 10% 12% 14%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Quarter since the last quarter before policy change
youngest kid 12-‐15, 2008Q3 youngest kid10-‐11, 2009Q3
youngest kid 7-‐9, 2010Q3 youngest kid 5-‐6, 2011Q3
Impact of SPA increase for women on employment
30%
35%
40%
45%
50%
55%
60%
Counterfactual male 60-‐64 employment rate
LFS employment rate 60-‐64 men
Counterfactual female 60-‐64 employment rate
LFS employment rate 60-‐64 year old women
Note: counterfactual employment rates are estimated. See Cribb et al (2013) “Incentives, shocks or signals: labour supply effects of increasing the female state pension age in the UK”
Employment rate for older workers: women aged 60-‐64
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
45.0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Euro area (13 countries)
Germany (unHl 1990 former territory of the FRG)
Spain
France
Italy
United Kingdom
Employment rate for older workers: men aged 65-‐69
0.0
5.0
10.0
15.0
20.0
25.0
30.0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Euro area (13 countries)
Germany (unHl 1990 former territory of the FRG)
Spain
France
Italy
United Kingdom
Employment rate for older workers: men aged 60-‐64
0.0
10.0
20.0
30.0
40.0
50.0
60.0
70.0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Euro area (13 countries)
Germany (unHl 1990 former territory of the FRG)
Spain
France
Italy
United Kingdom
For the young employment fell back.... Employment rate: men aged 25-29
60.0
65.0
70.0
75.0
80.0
85.0
90.0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Euro area (13 countries)
Germany (unHl 1990 former territory of the FRG)
Spain
France
Italy
United Kingdom
Male employment by age -‐ UK
Source: Blundell, Green and Jin (2014)
Employment with at least BA: men
Source: Blundell, Green and Jin (2014)
Employment, le^ school at 16: men
Source: Blundell, Green and Jin (2014)
Employment, le^ school at 16: women
Source: Blundell, Green and Jin (2014)
Hours, le^ school at 16: men
Source: Blundell, Green and Jin (2014)
2. Incomes • Leading up the recession:
• income growth had slowed in early 2000s. • pensioners/ working-‐age childless were doing relaHvely well/badly
• During recession and immediately a^erwards: • real earnings for those in work fell • employment rates fell for low skilled young adults but not for older ones • benefits/tax credit incomes were robust.
– As a result (…..drawing from IFS Living Standards Report): a) income inequality fell (despite rise in earnings inequality among workers) b) low educated young adults did worst; pensioners did best
© InsHtute for Fiscal Studies
-‐1.0%
-‐0.5%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
10 20 30 40 50 60 70 80 90 Average an
nual re
al income chan
ge Parents and children
Working-age without children
Pensioners
Percentile point
1996-‐97 to 2011-‐12: pensioners did relaHvely well; working-‐age childless relaHvely badly
Notes and source: see Figure 5.2 of Living Standards, Poverty and Inequality in the UK: 2013
Income trends not uniform: slower growth from early 2000s
100
105
110
115
120
125
1997–98=100
Mean income Median Income
© InsHtute for Fiscal Studies
Source: Family Resources Survey, various years
100
105
110
115
120
125
1997–98=100
Mean income Median Income
© InsHtute for Fiscal Studies
Source: Family Resources Survey, various years
Income trends not uniform: slower growth from early 2000s
... and large falls a^er the recession
100
105
110
115
120
125
1997–98=100
Mean income Median Income
© InsHtute for Fiscal Studies
Source: Family Resources Survey, various years
© InsHtute for Fiscal Studies
Source: Table 2.3 of Living Standards, Poverty and Inequality: 2014 Notes: This is a very slightly different sample to the overall income staHsHcs. Households with negaHve incomes are dropped. This makes a small difference to falls in income
Income sources, 2007–08 to 2009–10: steady income growth due to benefits/tax credits
-‐2 -‐1.5 -‐1 -‐0.5 0 0.5 1 1.5 2 2.5 3
Total income
Taxes and other deducHons
Other
Self-‐employment income
Savings and private pension income
Benefits and tax credits
Earnings
ContribuHon to income growth between 2007–08 to 2009–10 (in percentage points)
0.1
0.4
0.2
0.2
2.4
2.2
-‐0.7
© InsHtute for Fiscal Studies
Source: Table 2.3 of Living Standards, Poverty and Inequality: 2014 Notes: This is a very slightly different sample to the overall income staHsHcs. Households with negaHve incomes are dropped. This makes a small difference to falls in income
Income sources, 2009–10 to 2012–13: large income falls due to falling earnings
-‐10 -‐8 -‐6 -‐4 -‐2 0 2 4
Total income
Taxes and other deducHons
Other
Self-‐employment income
Savings and private pension income
Benefits and tax credits
Earnings
ContribuHon to income growth between 2009–10 to 2012–13 (in percentage points)
-‐8.1
-‐1.7
-‐0.2
3.2
-‐8.6
-‐1.0
-‐0.8
© InsHtute for Fiscal Studies
Source: Table 2.3 of Living Standards, Poverty and Inequality: 2014 Notes: This is a very slightly different sample to the overall income staHsHcs. Households with negaHve incomes are dropped. This makes a small difference to falls in income
Income sources: 2007–08 to 2012–13
-‐10 -‐8 -‐6 -‐4 -‐2 0 2 4 6
Total income
Taxes and other deducHons
Other
Self-‐employment income
Savings and private pension income
Benefits and tax credits
Earnings
ContribuHon to income growth between 2007–08 to 2011–12 (in percentage points)
-‐8.2
-‐1.4
0.1
3.5
-‐6.4
1.2
-‐1.5
-‐14.0% -‐13.0% -‐12.0% -‐11.0% -‐10.0% -‐9.0% -‐8.0% -‐7.0% -‐6.0% -‐5.0% -‐4.0% -‐3.0% -‐2.0% -‐1.0% 0.0%
10 20 30 40 50 60 70 80 90
Cumula4
ve re
al cha
nge Percentile point
Weekly earnings inequality (among workers) rose between 2007-‐08 to 2012-‐13...
Note: Excludes self-employment income Source: Family Resources Survey, various years
Note: towards bottom, much driven by falls in hours worked (not relatively larger fall in hourly wages)
...but net result was sHll a fall in income inequality
© InsHtute for Fiscal Studies
-12%
-8%
-4%
0%
4%
8%
10 20 30 40 50 60 70 80 90
Inco
me
chan
ge
2007–08 to 2009–10 2009–10 to 2012–13 2007–08 to 2012–13
Percentile point
Source: Family Resources Survey, various years
3. Consumption
• Expenditure falls have been deeper than in previous recessions. – note that the start of the fall is coincident with the fall in GDP (not
income).
• Unusually expenditure on consumer nondurables has fallen most – especially among the young families with children and to some extent
among the middle aged. Less for the old.
• This points to an expectaHon of a permanent/persistent fall in living standards, – especially among the young and middle-‐aged.
• The very recent growth is dominated by durable expenditure increases.
© InsHtute for Fiscal Studies
ConsumpHon Growth
© InsHtute for Fiscal Studies
80
85
90
95
100
105
2005Q1
2005Q3
2006Q1
2006Q3
2007Q1
2007Q3
2008Q1
2008Q3
2009Q1
2009Q3
2010Q1
2010Q3
2011Q1
2011Q3
2012Q1
2012Q3
2013Q1
Private consum
p4on
(2008Q
1 = 100) EU28
Spain
Italy
United Kingdom
Non-‐ (and semi) durables; UK recessions
© InsHtute for Fiscal Studies
90
95
100
105
110
115
120
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Qua
rter before recession = 100
Quarters since start of recession
1980
1990
2008
ConsumpHon DistribuHon by Year -‐ UK
Notes: Non-durables and Services; FES Source: Blundell and O’Dea (2014).
0
100
200
300
400
500
600
700
800
900 1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
Year
mean
p90
p75
p50
p25
p10
Consumption started falling back earlier than income…….
© InsHtute for Fiscal Studies
Percentage Change in Food Expenditure: UK
Notes: Understanding Society Source: Blundell and Etheridge, 2014
-‐0.1
-‐0.08
-‐0.06
-‐0.04
-‐0.02
0
0.02
0.04
0.06
0.08
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97 2012-‐2009
2012-‐2010
2010-‐2009
© InsHtute for Fiscal Studies
Percentage Change in Food Expenditure: 2010-‐2012 UK
-‐0.12
-‐0.1
-‐0.08
-‐0.06
-‐0.04
-‐0.02
0
0.02
0.04
1 4 7 10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73 76 79 82 85 88 91 94 97
Change in Food Expenditure by Cen4le and HH type: 2010 to 2012
All HHs
Pensioners
w/o children
with children
Notes: Understanding Society Source: Blundell and Etheridge, 2014
Prospects • Younger workers and families are acHng as if they expect a long-‐run fall
in relaHve living standards, and quite far up the distribuHon.
– evidence from consumpHon, savings and labour supply. • The number of rouHne jobs near the middle of the earnings distribuHon
has declined steadily
– more jobs are now professional or managerial – ‘successful’ absorpHon of the higher numbers with degrees.
Employment shares of occupation groups
© InsHtute for Fiscal Studies
Source: Blundell, Green and Jin (2014) Notes: the discontinuities in occupation classification in 2001 and 2011 have been addressed in the following way. For the conversion of SOC 1990 to SOC 2000, we look at individuals who were surveyed in 2000Q4 and 2001Q2 and stayed with the same employer and infer the transition matrix from this group. For the conversion of SOC2010 to SOC 2000, we used the ONS two-way tabulation of the LFS 2007 Q1 sample by the major occupation groups under the two SOC systems.
Prospects • Younger workers and families are acHng as if they expect a long-‐run fall
in relaHve living standards, quite far up the distribuHon. – evidence from consumpHon, savings and labour supply.
• The number of rouHne jobs near the middle of the earnings distribuHon has declined steadily – more jobs are now professional or managerial, ‘successful’ absorpHon of the higher
numbers with degrees.
• Suggests longer term earnings growth will mostly come from high-‐skilled occupaHons, enhancing the top of the income distribuHon.
• Lirle evidence of earnings progression for lower skilled. • Benefit changes have play a central role for low skilled
Source: Blundell, Dias, Meghir and Shaw (2013)
1.61.8
22.2
2.42.6
log w
age
20 30 40 50age
secondary further higher
Wage profiles by education and age – UK Women
Benefit changes are also important drivers SimulaHons up to 2015–16
© InsHtute for Fiscal Studies
-8%
-6%
-4%
-2%
0%
2%
4%
10 20 30 40 50 60 70 80 90
Inco
me
chan
ge
2007–08 to 2012–13
2012–13 to 2015–16
Percentile point
Note: Calculations using the Family Resources Survey 2007–08 and 2012–13 and TAXBEN, the IFS tax and benefit microsimulation model. Figure is an update of that in Brewer et. al. (2013), Fiscal Studies, Vol. 34, No. 2, pp. 179–201. See also Andy Hood’s presentation!
Benefit changes are also important drivers SimulaHons up to 2015–16
© InsHtute for Fiscal Studies
-8%
-6%
-4%
-2%
0%
2%
4%
10 20 30 40 50 60 70 80 90
Inco
me
chan
ge
2007–08 to 2012–13
2012–13 to 2015–16
2007–08 to 2015–16
Percentile point
Note: Calculations using the Family Resources Survey 2007–08 and 2012–13 and TAXBEN, the IFS tax and benefit microsimulation model. Figure is an update of that in Brewer et. al. (2013), Fiscal Studies, Vol. 34, No. 2, pp. 179–201. See also Andy Hood’s presentation!
Implications for Inequality • Persistent changes
– evidence from distribuHon of consumpHon, and labour supply responses, implies strong persistent/uninsurable shocks for certain key birth cohorts.
• Longer term earnings growth will mostly come from high-‐skilled occupaHons, – enhancing the top deciles of the income distribuHon.
• Lirle progression and low producHvity for the low skilled part of the distribuHon – incomes at the borom will only be maintained through family labour supply and the
benefit/tax-‐credit system (and younger adults staying at home).
• Studying the the relaHonship between inequality in income, consumpHon, wages and earnings brings structure to the key features of inequality and living standards – consumpHon falls much more than in past recessions, even for necessiHes, – labour supply of women and older workers is maintained, even enhanced, – for low skilled with few assets, family labour supply, family structure and welfare are
key insurance mechanisms.
• Providing a picture of the dynamics of inequality and living standards over the austerity period and a sense of the likely evoluHon in coming years.
• SHll need to do more on this with Understanding Society…..
© Institute for Fiscal Studies
That’s all!
Income Shocks, Consumption and Labour Supply: Understanding Inequality Dynamics
Richard Blundell University College London and Institute for Fiscal Studies
Thanks to Ben Etheridge, Andy Hood, Wenchao Jin, Robert Joyce, Cormac
O’Dea, Luigi Pistaferri, Ian Preston, and Itay Suporta
This research is funded by the ERSC Centre (CPP) at IFS.
MISOC Workshop September 15th 2014
© Institute for Fiscal Studies
Extra Slides
Income Shocks, Consumption and Labour Supply: Understanding Inequality Dynamics
Richard Blundell University College London and Institute for Fiscal Studies
Thanks to Ben Etheridge, Andy Hood, Wenchao Jin, Robert Joyce, Cormac
O’Dea, Luigi Pistaferri, Ian Preston, and Itay Suporta
This research is funded by the ERSC Centre (CPP) at IFS.
MISOC Workshop September 15th 2014
Employment and self-‐employment rate of older people
-‐10%
-‐8%
-‐6%
-‐4%
-‐2%
0%
2%
4%
2001Q1
2001Q3
2002Q1
2002Q3
2003Q1
2003Q3
2004Q1
2004Q3
2005Q1
2005Q3
2006Q1
2006Q3
2007Q1
2007Q3
2008Q1
2008Q3
2009Q1
2009Q3
2010Q1
2010Q3
2011Q1
2011Q3
2012Q1
2012Q3
2007Q4 = 0
employment rate 60-‐74 male self-‐employment rate 60-‐74 male
employment rate 60-‐74 female self-‐employment rate 60-‐74 female
(RPI deflated) Real wages by occupation group since 1993
© InsHtute for Fiscal Studies
Note: the low-skilled wage would end up around the 1993 level if we use CPI instead of RPI. Each log wage series is normalized to 0 in 1993.
NEET rate among young people
0%
10%
20%
30%
40%
50%
60%
18-‐21, GCSEs or above
18-‐21, no GCSEs or equilvants
22-‐24, degree or above
22-‐24, GCSEs, no degrees
22-‐24, no GCSEs
Median male real hourly wages by educaHon
Source: Blundell, Green and Jin (2014)
Note: Average real hourly wages have fallen back – this picture stands even a^er adjusHng for selecHon.
Adjusted Median Wage by Year, Men
Source: Blundell, Green and Jin (2014)
In contrast to previous recessions, real output per hour has at best been stagnant since 2008
90
95
100
105
110
115
120
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Quarter since the labelled one
Real output per hour
1979Q4 1990Q2 2008Q1 Linear trend (90Q2-‐08Q1)
Source: Blundell, Crawford and Jin (2014)
What have we learned? • Changes in the labour market are a key factor behind the changes in overall
living standards and inequality – especially for younger families/singles
• Real wages and employment tell much of the story – real wages have fallen across the distribuHon, – large increases in entry cohorts with BA have maintained their BA premium, even in the
recession but they have also suffered the overall decline in real wages,
– the poor employment and earnings performance of young low educated.
• Benefit changes are also important drivers – key impact in the near future too …
© InsHtute for Fiscal Studies
Prospects - summary • Younger workers and families are acHng as if they expect a long-‐run fall
in relaHve living standards, quite far up the distribuHon. – evidence from consumpHon, savings and labour supply.
• Real wages (and producHvity) have been slow to pick up – expect the parern of lower real wages to conHnue, but with fairly buoyant
employment due to increased supply.
• Most falls in real earnings have happened (but low/no real growth) – fiscal contracHon implies large benefit cuts in near future.
• The number of rouHne jobs near the middle of the earnings distribuHon has declined steadily – more jobs are now professional or managerial, ‘successful’ absorpHon of the higher
numbers with degrees.
• Suggests longer term earnings growth will mostly come from high-‐skilled occupaHons, enhancing the top of the income distribuHon.
• Lirle evidence of earnings progression for lower skilled.
Prospects - II • But sHll much to do in focussing on older workers in general, on return
to work for parents/mothers, and on entry into work.
• There are sHll some potenHal big gains here, – for example, as (higher skilled) women age in the workforce.
• Tax/welfare reforms to enhance earnings (from Mirrlees): – refocus incenHves towards transiHon to work, return to work for lower
skilled mothers and on enhancing incenHves among older workers.
• ProducHvity is sHll the key, – (financial) capital misallocaHon and potenHal investment returns.
• Human capital and ‘on the job’ wage/producHvity complementarity – note the relaHve importance of mismatch of entry skills in the recession.
• ProducHvity and wages are closely related but note recent growth in the wedge between labour costs and hourly wages, – the growing importance of pensions and NI in the UK, – what will the trends in this wedge look like?
Employer contribuHons to pension funds – in constant prices terms
Source: Office for National Statistics Notes: Data for Q4 2012 is not yet published so has been estimated based on Q4 2011 to Q3 2012 data
The growth in durable and non-‐durable real expenditure
© InsHtute for Fiscal Studies
80
85
90
95
100
105
110
115
2008 Q1
2008 Q2
2008 Q3
2008 Q4
2009 Q1
2009 Q2
2009 Q3
2009 Q4
2010 Q1
2010 Q2
2010 Q3
2010 Q4
2011 Q1
2011 Q2
2011 Q3
2011 Q4
2012 Q1
2012 Q2
2012 Q3
2012 Q4
2013 Q1
2013 Q2
Non-‐durable Durables
Employment rate by highest qualificaHons achieved, 16-‐59 year olds
© InsHtute for Fiscal Studies
Note: level 2 is GCSE grade C or above.
50%
55%
60%
65%
70%
75%
80%
85%
90%
95%
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
degree or above
level 2 and 3
level 1 or below
ParHcipaHon rates over Hme
0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0%
100.0%
23-64, male 23-64, female 65-84, male 65-84, female
• In this notation, the transmission parameters φ and ψ subsume π and γ from the self-insurance model
• This provides the key linkage between the evolution of distribution of consumption and the evolution of income
• It describes how the consumption distribution updates to income shocks
ln 'it it it t it t it itC Z ϑ φζ ψ ε ξΔ ≈ Γ +Δ + + +
Partial insurance coefficient w.r.t. permanent shocks, 0≤φ ≤1
Excess sensitivity coefficient w.r.t. transitory shocks, 0≤ψ≤1
Need to generalise to account for additional ‘insurance’ mechanisms and excess sensitivity
Consumption dynamics (2)
% facing nominal wage freeze in the coming year by current wage quintile
Source: New Earnings Survey Panel Dataset 1975-‐2012. There are 20,000-‐30,000 observaHons underlying each data point.
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
Lowest-‐paid 20% 2nd quinHle 3rd quinHle
4th quinHle Highest-‐paid 20%
More nominal pay freezes in the absence of collecHve agreement, 2008-‐11
Change to the proporHon of 23-‐64-‐year-‐olds who are unemployed by recession
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
0 1 2 3 4 5
male, from 1979
female, from 1979
male, from 1990
female, from 1990
male from 2008
female, from 2008
Source: LFS every year. No data point for year 1980, 1982. Quarter 2 is used for years since 1992.
Median male real hourly wages by educaHon (exl. Im.)
Source: Blundell, Green and Jin (2014)
Can changes in the composiHon of the workforce explain the fall in real hourly wages?
• Short answer: no – unlike, to an extent, in other countries.
• As in previous recessions, the composiHon of the workforce shi^ed towards more producHve workers, – hence we would have expected the average real wage to increase
(other things being equal).
• What has been different in this recession is that the returns to those characterisHcs have fallen substanHally.
• This is true even amongst workers who keep their jobs, who have experienced nominal wage freezes/real wage falls, – May have been facilitated by the reduced power of labour market
insHtuHons (e.g. unions) since previous recessions.
ProporHonate Fall in Each Component
© InsHtute for Fiscal Studies
65
70
75
80
85
90
95
100
105
110
Chan
ge since 2008 Q1 (2008 Q1 = 100)
Govt. purchases
Consumer durables
Nondurable consumpHon
Corporate investment
House Prices by Region
© InsHtute for Fiscal Studies
Components of GDP
© InsHtute for Fiscal Studies
-‐20
-‐15
-‐10
-‐5
0
5
10
15
Chan
ge since 2008 Q1 (£b) Net exports
Govt. purchases
Consumer durables
Nondurable consumpHon
Corporate investment
-‐1.0%
-‐0.5%
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
3.5%
4.0%
10 20 30 40 50 60 70 80 90 Average an
nual re
al income chan
ge Parents and children
Working-age without children
Pensioners
Percentile point
...and on an a^er-‐housing-‐cost (AHC) basis
Notes and source: see Figure 5.2 of Living Standards, Poverty and Inequality in the UK: 2013
Median income changes by age (BHC, GB)
Source: Family Resources Survey, various years
-4%
-3%
-2%
-1%
0%
1%
2%
3%
4%
0s 10s 20s 30s 40s 50s 60s 70s
Aver
age
annu
al g
row
th
2007–08 to 2012–13 2002–03 to 2007–08
Age
Average real hourly wages have also been stagnant
95
100
105
110
115
120
0 1 2 3 4 5 6
Real hou
rly wage (first y
ear o
f recession = 100)
Years since the year in which the recession began
Average real male hourly wage (using GDP deflator)
1979 1990 2008 Linear trend 1990-‐2008
Average real hourly wage by age group (RPI deflated)
0
2
4
6
8
10
12
14
16
18
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
16-‐17
18-‐25
26-‐35
36-‐45
46-‐55
56-‐64
65+
Employment and labour market parHcipaHon -‐ employment rate of 23-‐64-‐year-‐olds by recession
-10.0%
-8.0%
-6.0%
-4.0%
-2.0%
0.0%
2.0%
0 1 2 3 4 5
Year since the start of recession
male, from 1979 female, from 1979 male, from 1990 female, from 1990 male from 2008 female, from 2008
Source: LFS every year. No data point for year 1980, 1982. Quarter 2 is used for years since 1992.
Employment status of 16-‐22-‐year-‐olds
0%
10%
20%
30%
40%
50%
60%
70%
80% % in work % unemployed
Source: LFS 2nd
quarter every year
Male employment rate by age group over Hme
Source: Blundell, Green and Jin (2014)
3. Consumption
• Expenditure falls have been deeper than in previous recessions. – Note that the start of the fall is coincident with the fall in GDP (not income).
• Unusually expenditure on consumer nondurables has fallen most – Especially among the young families with children and to some extent among the
middle aged. Less for the old.
• Homeowners have made the largest cuts. – In past recessions there was lirle difference between owners and renters.
• Saving raHos are lower than during the early 1980s and early 1990s – but haven risen dramaHcally since 2008 (and pension contribuHons higher).
• The data points to an expectaHon of a permanent/persistent fall in living standards, – especially among the young and middle-‐aged.
• The very recent growth is dominated by durable expenditure increases.
© InsHtute for Fiscal Studies