ZEW LUNCH DEBATE

ECONOMIC RECOVERY IN EUROPE

INEQUALITY PATTERNS AFTER DOWNTURNS:

WHAT LESSONS FOR SOCIAL POLICIES?

Herwig Immervoll

Social Policy Division, OECD

Herwig.IMMERVOLL@oecd.org

Brussels, September 25, 2014

2

• Economic & fiscal crisis is also a social crisis:

much greater need for support

• Crisis preparedness instead of ad-hoc emergency

measures. ‘Crisis proofing’ should to be central design

feature of social protection, during all stages of

economic cycle

• ‘Active’ social policies reduce long-term costs of crises

by adapting support to changing economic and social

conditions

Summary: Main messages

3

Economic output has begun recovery everywhere

but employment and wages have not

Notes: Data are annual and changes are in real terms. Countries without drops in annual GDP between 2007-9 are excluded. Data in shaded area are projected.

Source: Society at a Glance 2014, using OECD Economic Outlook No. 93.

80

85

90

95

100

105

110

115

p-4 p-3 p-2 p-1 peak p+1 p+2 p+3 p+4 p+5 p+6

pre-crises peak=100

years since GDP peaked

GDP, median country Wage bill, median country

4

Economic output has begun recovery everywhere

but employment and wages have not

80

85

90

95

100

105

110

115

p-4 p-3 p-2 p-1 peak p+1 p+2 p+3 p+4 p+5 p+6

pre-crises peak=100

years since GDP peaked

GDP, median country Wage bill, median country

GDP "high-growth" countries Wage bill "high-growth" countries

Notes: Data are annual and changes are in real terms. Countries without drops in annual GDP between 2007-9 are excluded. Data in shaded area are projected.

Source: Society at a Glance 2014, using OECD Economic Outlook No. 93.

5

Economic output has begun recovery everywhere

but employment and wages have not

80

85

90

95

100

105

110

115

p-4 p-3 p-2 p-1 peak p+1 p+2 p+3 p+4 p+5 p+6

pre-crises peak=100

years since GDP peaked

GDP, median country Wage bill, median country

GDP "high-growth" countries Wage bill "high-growth" countries

GDP "low-growth" countries Wage bill "low-growth" countries

Notes: Data are annual and changes are in real terms. Countries without drops in annual GDP between 2007-9 are excluded. Data in shaded area are projected.

Source: Society at a Glance 2014, using OECD Economic Outlook No. 93.

6

Previous crises: Income gaps widen in/after

recessions, often persist during recoveries

• Income gaps widen – and

recoveries often fail to close them

• At bottom: incomes fall rapidly

during & after recessions, remain

low for long time

• At top: incomes often continue to

rise (at reduced pace)

• Any narrowing of gaps after

recessions does not last long

enough to close gap

• Longer time series for several other

countries similar

• Data for recent crisis point in similar

direction as historic trends

020406080100120140160 Negative-growth years Low-growth years

median incomes low incomes

high incomes

Household market incomes, working-age in real terms, earliest data point = 100

0

20

40

60

80

100

120

140

160

180

200

UnitedKingdom

“Low” and “high” incomes refer to the 10th and 90th percentiles.

Source: Immervoll and Richardson (2011), using LIS data.

7

Some signs that similar patterns

in recent downturns

0

20

40

60

80

100

120

140

160United States

0

20

40

60

80

100

120

140

160Finland

0

20

40

60

80

100

120

140

160Canada

0

20

40

60

80

100

120

140

160

180

200

220United Kingdom

0

20

40

60

80

100

120

140

160Denmark

8

Declining market incomes at the bottom:

Challenges & scope for social policy

0

20

40

60

80

100

120

140

160

Finland

9

Declining market incomes at the bottom:

Challenges & scope for social policy

0

20

40

60

80

100

120

140

160

Finland

Cushioning income losses

10

Declining market incomes at the bottom:

Challenges & scope for social policy

0

20

40

60

80

100

120

140

160

Finland

Cushioning income losses Supporting self-sufficiency

11

Policy challenge #1

Effective safety nets

12

Very large increases in the number of workless

households are a major test for social policies Shares of adults living in workless households, %

Note: Households are defined as “workless” if all household members are either unemployed or labour-market inactive.

“Adults” refers to the population aged 15-64.

2013 instead of 2012 for the United States.

Source: OECD estimates based on the European Union Labour Force Survey and the United States Current Population Survey.

0

3

6

9

12

15

18

21

%

2012 (↗) 2007

13

Struggle to meet basic needs, negative

consequences for current & future well-being Share of people feeling they cannot afford food, %

Note: Share of “yes” responses to the question “Have there been times in the past 12 months when you did not have enough money to buy food that you or your family needed?”.

Source: GallupWorld Poll, www.gallup.com/strategicconsulting/en-us/worldpoll.aspx.

0

5

10

15

20

25

30

35

40

45

502011/12 (↘) 2006/07

14

Limited responsiveness of assistance benefits Recipient numbers, OECD total, 2007=100

Source: OECD Social Benefit Recipients database, forthcoming.

90

100

110

120

130

140

150

160

170

180

190

2007 2008 2009 2010

Unemployment insurance

Unemployment assistance

Social assistance

Disability pension

Old age pension

15

• Social insurance where affordable

• Complemented by effective safety nets:

– responsive to downturn, i.e. ready to scale up

– adequate resources for meeting basic needs

– ensure health and human capital is maintained

• Support families’ own capacity to adapt to

economic difficulties

Some key challenges for social

protection…

16

Policy challenge #2

Suitably targeted support

17

Job and income losses were not symmetric Change in shares of people without work, by age, sex and edu

weighted average, percentage-point change Q4 2007-Q4 2012

Source: OECD (2013), OECD Employment Outlook.

-5

-3

-1

1

3

5

7

9

11

13

- skill + - skill + - skill + - skill + - skill + - skill +

Men Women Men Women Men Women

Youth Prime Old

%

Inactivity Unemployment Non-employment

OECD total

18

Job and income losses were not symmetric Incomes mostly fell, but elderly either not affected, or less so

(annual % change 2007-2011 in real terms)

Source: OECD Income Distribution Database. www.oecd.org/social/inequality.htm.

- 12

- 10

- 8

- 6

- 4

- 2

0

2

4

6

Total (↗) Young (18-25 y.o)

19

Job and income losses were not symmetric Incomes mostly fell, but elderly either not affected, or less so

(annual % change 2007-2011 in real terms)

Source: OECD Income Distribution Database. www.oecd.org/social/inequality.htm.

- 12

- 10

- 8

- 6

- 4

- 2

0

2

4

6

Total (↗) Young (18-25 y.o) Elderly (Over 65 y.o)

20

A longer-term trend:

Shift of poverty risks from elderly to youth Rates of relative poverty, by age group

Source: OECD Income Distribution Database. www.oecd.org/social/inequality.htm.

50

75

100

125

150

175

200

50

75

100

125

150

175

200

Below 18 18-25 26-40 41-50 51-65 66-75 Above 75

Total poverty rate = 100 OECD country average

Mid-1980s

Mid-1990s

2007

2011

21

When social transfers are highly targeted,

spending cuts are more likely to hurt the poor Cash transfers received by low- and high-income groups, % of average transfers in 2010

Reading note: In Portugal, the average total transfer payment received by low-income families (in the bottom 30% of the income distribution) is 71% of the average

payment across all families, and less than half of the average benefit payment received by high-income families, who receive 52% more than the average

family.

Notes: Reference year is 2009 for Hungary, Japan, New Zealand, Switzerland and Turkey. “Bottom 30%” and “top 30%” refer to average public transfers received

by decile groups 1 to 3 and 8 to 10, respectively. Decile groups are determined in relation to household disposable income after accounting for taxes and

transfers. All incomes and transfer amounts are adjusted for household size (see www.oecd.org/social/inequality). Transfers include all public social benefits.

Source: OECD Income Distribution Database, www.oecd.org/social/inequality .

0

50

100

150

200

250% Bottom 30% (↗) Top 30%

22

• Rebalance support for different groups

• Address structural factors that lead to

unequal burdens for different groups

– Social protection provisions that render support

largely inaccessible for some groups

– Labour-market inequality (“insider/outsider”)

Some key challenge for social

protection…

23

Policy challenge #3

Adaptive social policies

24

• Adaptive social policy structures – Out-of-work benefits: Good social & economic arguments for modifying benefit

provisions when labour-market conditions change substantially

– Active labour-market policies

– In-work benefits and other into-work support

• Adaptive social policy resources

Adaptive social policies

25

Automatic stabilisers: role in limiting

income losses among poor

Notes: The “Strength of automatic income stabilisers” is a coefficient that shows how changes in market income translate into changes in disposable income.

Source and details: Society at a Glance, 2014.

AUT

BEL DNK

EST

FIN

FRA

DEU

GRC

HUN IRL

ITA

NLD

PRT

SVN

ESP

SWE

GBR USA

-15

-13

-11

-9

-7

-5

-3

-1

1

3

5

0 0.2 0.4 0.6 0.8

Strength of automatic income stabilizers

disposable income for bottom 10%, annual change 2007-10

26

Rising social spending and social needs,

but decreasing fiscal space Fiscal space narrowing, especially in high-unemployment countries

Note: Averages for 2007/8 and 2011/12 are used as the timing of the downturn and the beginning of any fiscal consolidation efforts varied across countries.

“Consolidation effort”: change in underlying primary government balance, percentage points of GDP. Correlation: 0.69, excl. GRC, IRL, PRT and ESP: 0.46

Source: OECD (2013), Economic Outlook Database, No. 93, June, www.oecd.org/economy/outlook/economicoutlook.htm.

AUSAUT

BELCAN

CHE

CZE

DEU

DNKEST

FIN

FRA

GBR

HUNISL

ISR

ITA

JPN

KOR

LUX NLD

NORNZL

POL

SVN

SWEUSA

8.96

10.39

-4

-2

0

2

4

6

8

10

-1 0 1 2 3 4 5

Chang

e in u

nem

plo

ym

ent ra

te,

2007/8

-2013

Consolidation effort, 2011/12-2014

IRLPRT

ESP GRC

8

14

20

2 3 4 5

27

Drivers of widening budget deficits Changes in revenues and transfer spending, 2007-10, in real terms

Notes: The vertical y axis is inverted (a positive number indicates an increase in social benefit expenditure and a worsening budget balance). Government

transfers: all cash social benefits paid by government. Government revenues: total tax and non-tax receipts of the general government sector (central

and sub-central) plus social security contributions.

AUS

AUT

BEL

CAN

CZE

DNK

FIN

FRA

DEU

GRC

HUN

ISL

IRL

ISR

ITA

JPN

KOR

LUX

NLD

NZL NOR

POL PRT

SVK

SVN

ESP

SWE

CHE

GBR

USA

0

2

4

6

-10 -8 -6 -4 -2 0 2

Cash

go

vern

me

nt

tran

sfe

rs

(valu

es in r

evers

e o

rder)

Government revenues

Transfer spending is main

driver of deteriorating

budget balance

Lower revenues are main

driver of deteriorating

budget balance

28

• Counter-cyclical fiscal policies – Increase/maintain cyclicality of social spending

– Complement revenues; less variable tax bases

– Downturn: design and time spending cuts accordingly

– Upswing: build up reserves (earmarked

taxes/contributions; link support to labour-market

conditions)

• Continue structural reforms begun before

the crisis – Old-age pension

– Health

– Activation

Some key challenge for social

protection…

• ‘Efficiency’ of social polices: How do countries

adapt social policy efforts over the cycle? What

are effects on key social outcomes?

• Benefit dependence: Patterns and drivers of

benefit inflows, outflows, durations.

• Multi-dimensional “profiles” of populations with

labour-market difficulties: Inform targeting,

customising ‘active’ and ‘passive’ support

• Work incentives: When and for whom are they

relevant? New empirical work

How to inform and support this broad

agenda? Some activities at OECD

30

Thank you Contact: Herwig.Immervoll@oecd.org

OECD Directorate for Employment, Labour and Social Affairs, via www.oecd.org/els

Redistribution and working-age benefits, via www.oecd.org/social/benefitsandwagesindicators.htm

Employment policies and data, via www.oecd.org/els/emp

Income distribution and poverty, via www.oecd.org/social/inequality.htm

Society at a Glance, via www.oecd.org/els/societyataglance.htm

@OECD_Social

31

Additional slides

32

Political sustainability

Data are taken from Transatlantic Trends, an annual survey of public opinion by German Marshall Fund of the United States, Compagnia di San Paolo, Barrow

Cadbury Trust, Fundação Luso-Americana, BBVA Foundation, Communitas Foundation, and Swedish Ministry for Foreign Affairs. Polling for the 2013 results took

place in June and July by phone interview. In each country, the sample consists of approximately 1 000 randomly chosen men and women of 18 years of age and

older. The 95% confidence interval attributable to sampling and other random effects is no more than plus or minus three percentage points.

Decrease spending Maintain current levels of spending

Increase spending "don’t know”/refusal

0102030405060708090100 0 10 20 30 40 50 60 70 80 90 100

Portugal

France

Italy

United States

Slovak Republic

Germany

Spain

Poland

Netherlands

United Kingdom

Turkey

Sweden

1. “What should happen to

government spending?”

2. “What should happen to

welfare-state spending?”

33

Unemployment benefit generosity varies

enormously across countries Net replacement rate over a long unemployment spell

Year 1 2 3 4 5

Belgium 70 64 64 64 64

Ireland 62 62 62 62 62

Finland 64 62 35 35 35

France 67 67 29 29 29

Germany 67 35 35 35 35

Norway 73 73 17 17 17

Sweden 58 57 54 19 7

Iceland 60 56 56 7 7

Spain 68 65 23 23 2

Portugal 72 64 34 4 4

Denmark 74 74 9 9 9

United_Kingdom 33 31 31 31 31

Netherlands 73 61 5 5 5

Canada 58 16 16 16 16

United_States 50 34 0 0 0

Poland 45 8 8 8 8

Japan 47 5 5 5 5

Lithuania 31 8 8 8 8

Latvia 46 2 2 2 2

Greece 34 9 4 4 4

Italy 45 0 0 0 0

Year 1 2 3 4 5

34

Who receives benefits?

More or less strict gatekeeping Illustration: Strictness of entitlement conditions

Indicator on two aspects of entitlement conditions:

(a) employment/contribution requirements, (b) sanctions for “voluntary” unemployment Source: Venn (2012), “Eligibility Criteria for Unemployment Benefits: Quantitative Indicators for OECD and EU Countries”

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

100%

BG

R

AU

S

AU

T

CYP

DN

K

NZL

NO

R

CZ

E

FIN

JPN

GB

R

LTU

CH

E

SWE

DEU IS

R

FRA

HU

N

BE

L

IRL

SV

K

CA

N

GR

C

KOR

NLD

USA

PO

L

LUX

MLT

SLO

ESP

EST

RO

M ITA

TUR

PR

T

35

Who receives benefits?

Coverage has been falling in most countries % of ILO unemployed, selected countries

-5

5

15

25

35

45

55

65

75

851995 2005

Source: Immervoll and Richardson (2011) using European Labour Force Surveys and the US Current Population Survey.

36

Working-age social spending most-

frequently cited item in consolidation plans

Note: A questionnaire was submitted to countries in December 2011 to update The OECD 2010-11 report on fiscal consolidation strategies. Based on

country responses and publicly available sources, information was collected for 31 countries. Explanation: 70% of countries plan to cut welfare spending in

2012. Welfare includes working-age social transfers, such as unemployment benefits, social assistance, disability benefits, child benefits, etc. Pensions

comprise old-age pensions.

Source: OECD National Accounts, OECD (2012), “Restoring Public Finances, 2012 Update”, and OECD (2011), "Restoring Public Finances: Fiscal

Consolidation in OECD Countries".

Major programme measures in fiscal consolidation plans,

by domain, based on 2011 and 2012 surveys, OECD

Government spending by function,

OECD average, 2011

0

10

20

30

40

50

60

70

80

% o

f par

ticip

atin

g co

untr

ies

Plan 2012 Plan 2011Working-age

transfers

Health

Pensions

Other social protection

Other

Education

37

• Except if linked to unemployment levels (CHE, DNK, AUS)

Change in spending per unemployed, 2007-11

Active labour-market policies:

Falling resources per unemployed

-60

-50

-40

-30

-20

-10

0

10

20

Ireland Euro area OECD

%

Note: OECD is the weighted average of 32 OECD countries (excluding Iceland and Turkey). 2010 for Ireland.

Source: OECD estimates based on the OECD Labour Market Programmes and OECD Short-Term Labour Market Statistics Database.

Income support Re-employment assistance

38

• Unemployment insurance: strengthened, extensions now phased out. But some countries restricted durations or access early on (CZE, DNK, ESP, HUN, PRT)

Others moved later to strengthen benefits (ITA, PRT)

• Unemployment assistance: strengthened But cuts in HUN, PRT. Access remains very limited in GRC

• Social assistance safety nets: fewer changes, some strengthening But cutbacks or tighter eligibility in HUN, NLZ, PRT, UK

• Disability: stricter gate keeping, time limits, reassessments

• Family benefits: savings measures since 2010

• Old-age pensions: pre-crisis “structural” reform plans brought forward. New means-tested safety nets in CHL, FIN, GRC, MEX, enhanced in AUS, ESP

Some across-the-board cuts (GRC, freezes in AUT, GRC, ITA, PRT, SLO)

Key trends by policy area:

(1) Cash transfers

39

Thank you Contact: Herwig.Immervoll@oecd.org

OECD Directorate for Employment, Labour and Social Affairs, via www.oecd.org/els

Redistribution and working-age benefits, via www.oecd.org/social/benefitsandwagesindicators.htm

Employment policies and data, via www.oecd.org/els/emp

Income distribution and poverty, via www.oecd.org/social/inequality.htm

Society at a Glance, via www.oecd.org/els/societyataglance.htm

@OECD_Social