Lifelong Learning, Equality and Social Cohesion

Presentation at LLAKES Conference on ‘Lifelong Learning, Crisis and Social Change’

Senate House, London, 18th and 19th October 2012

Andy GreenDirector of ESRC-LLAKES Centre

Structure of Presentation

1. Social benefits of education at different levels

- to individuals- to communities (social capital)- to society (social cohesion)

2. Pathways for social effects of learning

3. The problem of educational inequality

4. Regimes of Social Cohesion, the Crisis and Education• What holds different societies together?• Recent trends and vulnerabilities in each regime

Individual Level Effects

Studies for various countries demonstrate that more educated

people tend to show higher levels of :

• Social and political trust• Civic and political engagement • Democratic values • Tolerance

and lower levels of violent crime.

(Nie et al., 1996; Stubager, 2008; Hagendoorn, 1999; Emler and Frazer, 1999;

Putnam, 2000). (Nie et al., 1996; Stubager, 2008; Hagendoorn, 1999; Emler and Frazer,

1999; Putnam, 2000; McMahon, 1999).

Some Findings from Analyses of UK Longitudinal Data

(Feinstein et al., 2003).

Compared with those educated to level 2 graduates are:

• 70-80% more likely to report excellent health

• 55% less likely to suffer depression (males)

• 3.5 times more likely to be a member of a voluntary association (Males: F= 2.5 x)

• 30% and 40% more likely to hold positive attitudes to race and gender equality

• 50% more likely to vote.

Education and Social Capital: Benefits to Communities

Education is also found to contribute to the social capital of groups and

communities where SC is defined as ‘features of social life – networks,

norms and trust – that enable to participants to act together more

effectively to pursue shared objectives’ (Putnam, 2006).

Putnam (2000) finds that more educed people are more likely to join

Associations’ make charitable donations and be politically engaged.

Repeated interactions in groups increases levels of trust and tolerance.

- Individuals thus benefit from enhanced networks - Neighbourhoods benefits from more co-operation and cohesion etc

Education and Social Cohesion

Social capital amongst individuals, families and local communities is not

the same thing as social cohesion at the country level.

Intra-group bonding does not always translate into inter-group harmony.

A country can have high levels of social capital in particular communities

but not be at all socially cohesive (eg Northern Ireland would be a good

example : see Schuller, Field et al, 2000).

It follows that:

Individual social benefits through increased learning do not necessarily

translate into societal effects or coincide with increased social cohesion.

LLAKES Research on Macro-Social Benefits

In our early research (Green, Preston and Janmaat, 2006) we found that

relationships pertaining at the individual level in some countries

disappear in macro-level, cross-country analysis.

• Social capital theorists argue that trust, civic engagement and tolerance go together at the individual level. However, they don’t co-vary across countries.

• Education enhances trust, tolerance and associational activity among individuals (in some countries). However, we found no relation across countries between adult skills and levels of trust, civic engagement and tolerance.

The Paradox of Levels

There are a number of reasons for this.

• The individual level effects are ‘relative’ or ‘positional’ ie one person’s social gain through improved learning outcomes will be another’s loss through relatively diminished skills.

• Other determininng factors at the national level overwhelm the statistical relation between education and social outcomes.

• Contexts: effects at the societal level are often indirect - ie they work through other factors which differ between societies.

LLAKES Research on LLL Social Benefits

• Uses mixed method multi-level approaches to understand relationships at different levels

• Draws on a range of different disciplines to understand the different mediating national contexts (labour market organisation; welfare systems etc)

• Examines both direct and indirect effects

Direct Effects Mediated by Contexts

Education can have direct effects on social

outcomes, it is argued, through raising cognitive

abilities and through socialisation into particular

sets of values and identities.

However, many of the direct effects are highly

influenced by (national) contexts.

Tolerance

Research for a number of countries shows that more educated people are more

tolerant (eg Putnam, 2000). It is argued that education can develop both cognitive

resources and values which protect against racial prejudice (Hagendorn, 1999).

However, there is no clear-cut relationship across countries between levels of education

and tolerance (Green, Preston and Janmaat, 2006) because other contexts, like the

political climate, vary and mediate the relationships.

Halman’s analysis (1994) of Eurobaromter data suggest that levels of tolerance in EU

countries vary according to the actual and perceived proportion of immigrants.

Jasinska-Kania analysis of EVS data (1999) shows that the impact of education on

racial tolerance is greater in countries with higher levels of immigrants (perhaps

because there are more circumstantially-driven racist attitudes that can be countered by

education).

Contextual Effects on Civic Participation

Various studies (eg Emler and Fraser, 1999) have shown a strong relationship at

the individual level between civic knowledge and civic activity. However, this

relationship does not necessarily hold at a national level.

The IEA Civic Education study of 14-year olds in 28 countries (Torney-Purta et al,

2001) found that levels of civic knowledge were relatively high in Finland, Norway,

Poland, Slovak Republic and Czech Republic. The context of the political changes

occurring in the transition countries no doubt contributed in the case of CEE countries.

Nordic countries scored low in support for different forms of political participation and

the Czech Republic low in support for non-conventional forms of civic engagement.

The Slovak Republic scored in high civic knowledge, but low in support for rights for

women and ethnic minorities (like Bulgaria, Estonia, Hungary, Latvia, Lithuania).

Contextual Effects on Education and Crime

McMahon (1999) used time lagged analysis of data for 78 countries (1956- 1995) and

found that rising rates of secondary education were associated with decreasing levels of

violent crime. Other contextual Factors are also important, however.

Junger-Tas (2000) finds that in countries such as England and Germany father absence

was associated with higher delinquency, but not on Nordic countries.

This is possibly due to different welfare arrangements between countries whereby single parent

families receive more support in Nordic states.

Similarly, whereas there was a relation between large peer groups and delinquency in

some countries, this was not the case in southern Europe where, arguably, these are

more common.

Positional Effects of Education on Political Engagement

Robert Nie et al. (2006), using OLS regression analysis on US time

series data, find that education has some absolute effects on political

engagement but the relative or ‘positional’ effects are stronger.

More educated people have more opportunity to achieve ‘network

centrality’ giving access to politicians, thus giving individuals an

incentive to participate. However, network centrality is a ‘zero-sum’

property - the gains for one individual will entail losses for others.

Thus while average education levels may be getting higher in North

America this does not necessarily lead to higher level of political

engagement.

Learning effects on social capital (joining, volunteering and engagement)  

  Learning Joining

volunteering civic engagement

   Cognitive resources (knowledge, skills etc)   Adapted from R. Nie

Status Network centrality

Which Effects are Absolute rather than Positional?

If individual social effects from learning are ‘absolute’ they are likely to aggregate into societal effects. If the are ‘relative’ or ‘positional’ they may not do so.

Recent research shows positional effects for a range of social outcomes.

• voter turnout (Burden, 2009; Tenn, 2007)

• political sophistication (Highton, 2009)

• democratic citizenship (Persson and Oscarsson, 2010).

Indirect Effects

Much of the influence that education has on social

outcomes is indirect – it works through something else.

LLAKES research suggests that often the most

powerful effects on social cohesion are distributional –

they depend on how the distribution of skills affects

the distribution of incomes and social status.

What matters most for social cohesion is less how much

education a country has, but how it is spread around.

Correlations between Adult Skills Distribution and Trust

We measured skills inequality using IALS cross-

country data on adult numerical skills, using the

‘test score ratio method’

Trust in other people is based on World Values

Survey Data.

POR

US

D

NW

UKB

CAN

PO

SZAU

IRLFIN

NLSW

DEN

0

10

20

30

40

50

60

70

1 1.1 1.2 1.3 1.4 1.5 1.6

Education Inequality

Gen

eral

Tru

st

NW

B

UK

CAN

IRL

NL

AUFIN

POSZ

POR

SWDEN

USA

D

1.00

1.10

1.20

1.30

1.40

1.50

1.60

20.00 25.00 30.00 35.00 40.00 45.00

Inco

me

ineq

ualit

y

Test score ratio

Inequality and Trust

Countries with more equal skills distributions have higher levels of trust.

This probably works partly through the effects of skills distribution on

income distribution, but the correlation exists independently of income

distribution. If the relationship is causal , it probably works both ways.

• Greater inequality of skills and incomes produces stress through creating high-stakes competition which reduces the capacity to trust in others.

• Inequalities in levels of education and skill increases CULTURAL DISTANCE between individuals and groups and makes trusting more difficult.

Over Time Analysis

Using time series data on education inequality, income inequality

and social cohesion measures over time (1960-1990) for

industrialised countries.

• Measure of educational inequality: Education Gini based computed from data on highest level of education

• Measure of unrest comprising riots, strikes and demonstrations.

• Measure of civil liberties based on freedom house scale.

-20

24

6S

tand

ardi

zed

valu

es o

f (u

nres

t1)

0 .2 .4 .6 .8edgini

Relationships

• Education inequality highly correlated with unrest but the relationship is non-linear. As education inequality rises ‘unrest‘ first drops slightly and then rises sharply.

• Educational inequality is generally negatively related to civil liberties but the relationship is again non-linear. As education inequalities rise, civil liberties first decline, then rise and then drop sharply.

Education Systems Properties and Civic Competences

Janmaat’s multi-level analysis of Cived data explored the effect of different

system and classroom characteristics on Civic competences.

Compared with comprehensive systems, selective education systems have:

• higher levels of social segregation across classrooms;

• greater disparities in civic knowledge and skills;

• larger peer effects on civic knowledge and skills - meaning that the latter are strongly affected by the social backgrounds and achievement levels of other students in the class.

(Janmaat , 2011).

Classroom Diversity and Values

Students who spend longer in mixed-ability classes are more likely to share basic values in areas such as tolerance and patriotism, regardless of their social own ethnic group (Janmaat & Mons 2011).

Ethnic diversity in the classroom seems to promote tolerance in some countries, but not in all.

In Germany and Sweden, native majority students tend to be more tolerant when in ethnically diverse classrooms.

In England, no such relationship was found. Furthermore, in English classrooms white students were less tolerant the better their minority ethnic peers performed in terms of civic knowledge and skills. This may again be related status and competition anxiety.

Macro Social Benefits Less Likely in Unequal Education Systems

LLL seems to be more successful in promoting social

cohesion in countries with more equal educational

outcomes.

• Nordic and East Asian countries tend to have relatively equal outcomes

• ‘Liberal’ and ‘Social market’ countries tend to have rather unequal outcomes.

Total Variance in Scores By Country Group: PISA 2000, 2009

48.5 46.2 44.83

35.5 34.631.75

0

10

20

30

40

50

US, UK Anglo Germanic E. Asia Nordic S.Europe

Score Point Difference Associated with One Unit on ESCS - Social Gradient

Liberal Social Market Southern Europe Eastern Europe East Asia Nordic0

5

10

15

20

25

30

35

40

45

44.6 43.85

31.25

40.33

3634.6

Score Point Difference Associeted with a One Point Increase in Student Background (ESCS) - 2009

Liberal Social Market Southern Europe Eastern Europe East Asia Nordic0

10

20

30

40

50

60

70

80

Percentage of Variation in Performance between Schools Explained by School ESCS - 2009

Adult Learning Not Mitigating Skills Inequalities in UK

• In Britain the well educated participate 1.6 times as much as the average person and the poorly educated participate only 0.3 times as much.

• In Canada, Denmark, Finland, Germany, the Netherlands, Norway, Sweden and the United States, the participation rates of both the high and low education groups are closer to the national mean (OECD, 2005 based on LFS data).

• In Britain the unemployed and inactive participate less than the national average.

• In Denmark, Germany, the Netherlands, Norway, Portugal, Spain and Sweden the unemployed have higher participation rates than the employed.

Population aged 25-29 by Highest Qualification Attained

55.0

19.3

50.2

5.7 8.917

65.8

12.2

59.5

44.6

28.0

14.9

37.6 34.8

46.5

0.0

20.0

40.0

60.0

80.0

100.0

UK (1998) Germany (1997) Singapore (1998) Korea (1998) Japan (1997)

Below Level 3Level 3Level 4 and above

Regimes of Social Cohesion

Historical and contemporary evidence suggests

that countries ‘hold together’ in different ways.

• Different historical traditions of thought on social

cohesion in different parts of the world.

• Different institutional arrangements support social cohesion.

Liberal Discourses

Liberal discourses tend to play down:

• The role of the state (in welfare and redistribution)• Equality• Shared values and identities (other than ‘core values’)

Emphasise importance of:

• Active civil society – at local level• Opportunity and individual liberty (‘core values’)• Tolerance

Republican Discourses

Republican discourses emphasise the state rather than civil society.

The state is seen to underpin social cohesion through:

• Providing welfare and social protection• Redistribution• Supervising conflict-mediating social partnership institutions• Promoting shared values and common national identity.

Different currents in republican thought variously stress equality of

opportunity or equality of outcomes as important pre-conditions for

social cohesion, but their role in social cohesion is often largely

symbolic.

Social Democratic Discourse

The social democratic discourse follows the republican discourse

in most of its essentials, except that here the stress on equality is

more profound.

• Like republican theory social democratic theory emphasises

both the role of the state and that of autonomous but state- sanctioned national civil society organisations

• Equality is seen as pre-condition of social solidarity.

• Common identity is highly valued.

Recent Research

Our recent research in LLAKES uses a wide range of

measures to test whether these different regimes

can be identified in contemporary societies.

The data:• Data on social attitudes from international surveys

(such as WVS and ISSP)• International administrative data

Component Tradition/regime Indicator(s)Indicators based on administrative data

Inequality Social Democratic (-)

Liberal (+)

Gini coefficient on household income

Wage regulation

Social Democratic (+)

Social Market (+)

Liberal (-)

Union coverage Centralization of wage bargaining

Employment protection

Liberal (-)

Social market (+)

Employment protection legislation 1998

State involvement

Liberal (-);

Social democratic (+);

Social market (+);

Public employment as percentage of total employment 2000

Welfare state Liberal (-);

Social democratic (+)

Public social expenditure as percentage of GDP 2000

Ethno-racial diversity

Liberal (+)

East-Asian (-)

Proportion of the population born abroad 2000

Crime / disorder

Liberal (+)

East Asian (-)

Social Market (-)

Homicide rate Violent crime 2000

Measures based on survey data

Social trust Social democratic (+)

Social Market (-)

East Asian (+)

Percentage saying most people can be trusted

Value diversity Social market (-)

East Asian (-)

Liberal (+)

Composite indicator representing the dispersion of opinions

Active civic participation Liberal (+)

East Asian (-)

Number of different voluntary organizations worked for

Passive participation in nationwide organizations

Social market (+)

Social democratic (+)

East Asian (-)

Number of different organizations belonging to

Freedom vs equality Liberal (+);

Social market (-);

Social democratic (-)

Freedom or equality more important; percentage preferring freedom

Merit vs equality Liberal (+);

Social market (+);

Social democratic (-)

Pay according to performance

Ethnocultural versus civic identities

Romantic conservative (+); East Asian (+); Liberal (-)

Strength of cultural relative to political conceptions of national identity

Ethnic tolerance Liberal (+); Romantic conservative (-); East Asian (-)

Xenophobia index; average (inverse indicator) Percentage not mentioning minding foreigners as neighbours

Social hierarchy East Asian (+);

Social market (+)

Percentage saying one should always love and respect one’s parents

Gender equality East Asian (-)

Social market (-)

Social democratic (+)

Liberal (+)

Percentage disagreeing that in times of scarcity men have more right to a job than women

Results

The statistical analysis uses :

• Correlations and scatter plots• Cluster analysis• Factor Analysis• Composite indicators and indexes.

Different regimes of social cohesion can be readily identified.

On all the tests countries and their social cohesion

characteristics cluster very much as the theory would

suggest.

Liberal Social Democratic Social Market East Asian

Country Score Country Score Country Score Country Score

16.81 SWE 15.90 AU 5.59 KOR 11.66

CAN 9.24 DEN 10.76 POR 3.12 JAP 9.10GB 4.43 NL 8.15 GER 3.05 CZE 3.37IRE -.14 FIN 7.42 FRA 2.27 POL 2.65GER -.74 B 3.11 ITA 1.82 ITA 2.34NL -1.93 AU .81 B .83 SP 2.02AU -2.05 GER .28 SWE .45 POR 1.97DEN -2.13 IRE .19 FIN -.37 SLV 1.21SP -2.27 SP -.42 NL -.59 GER -.12ITA -2.49 GB -.80 SP -1.74 AU -.52POR -2.86 FRA -1.10 DEN -2.84 IRE -.89FRA -3.96 CAN -2.62 IRE -3.14 FRA -1.35FIN -4.48 ITA -2.92 GB -5.54 FIN -2.00SWE -5.49 -3.26 CAN -6.76 GB -2.03B -6.08 POR -5.39 -11.33 NL -2.49

B -3.40DEN -3.69CAN -4.23SWE -7.24 -8.13

Rank order of countries on the four indexes

Trends

Trends in Social Trust

1981 1990 2000 200520

25

30

35

40

45

50

55

Liberal

Social Market

Southern European

East Asian

Trends in Political Trust

1981 1990 2000 200520

25

30

35

40

45

50

55

60

65

Social Democratic

Southern European

Social Market

Liberal

Current Vulnerabilities in Each Regime

Each regime of social cohesion is currently vulnerable at

the points most essential to its model.

• The Republican Regime has traditionally relied on widely shared common values. This is increasingly challenged by cultural diversity.

• The Social Democratic Regime relies heavily on its universalist welfare state. This is challenged by globalisation and immigration.

• The Liberal Regime relies on opportunity and the belief in meritocratic rewards to hold the together. This is challenged by rising inequality and declining social mobility (in UK and the US) particularly.

UK: The Atrophy of Core Beliefs in Meritocracy

Traditionally people in Britain are relatively tolerant of

inequality. But there is a large and probably growing gap

between people’s high expectations of meritocracy and

what they perceive to be the case.

Like people in Nordic Countries people are much more

likely than in most countries to say that effort

rather than need should determine pay.

But they are much less likely to perceive that

opportunities are in fact equal.

Country Hard work / children to

provide for *(ISSP 2009)

Country Large income dofferences acceptable to reward talents and effort (ESS 2008)

Australia 56.4 Greece 74.7

New Zealand 54.7 Denmark 66.7

Norway 51.6 Great Britain 63.9

Sweden 47.4 Germany 60.1

Great Britain 47.0 Netherlands 57.7

Finland 44.2 Switzerland 56.4

USA 44.2 Belgium 55.8

Japan 39.8 Cyprus 55.2

Iceland 39.2 Israel 54.5Portugal 35.4 Spain 52.9South Korea 30.3 Norway 52.6

Slovenia 28.7 France 51.5

Denmark 28.2 Sweden 49.0

Austria 22.9 Portugal 48.9

Switzerland 21.9 Slovenia 36.6

France 19.9 Finland 27.5

Belgium 19.4Spain 16.2Germany 10.8Israel 5.7

Country Only the rich can afford the cost of attending university (ISSP 2009)

Disagree

Country People have the same chances to enter university, regardless of their gender,

ethnicity or social background (ISSP 2009)

Norway 85.6 Germany 44.3

Denmark 83.9 France 41.6

Finland 80.9 Portugal 38.8

Iceland 73.9 Spain 29.1

New Zealand 66.6 Great Britain 28.6

Spain 66.5 Austria 26.8

Austria 66.4 Australia 25.2

Sweden 64.7 South Korea 23.3

Switzerland 64.6 USA 23.3

USA 61.4 Belgium 22

Belgium 52.4 Denmark 21.2

Australia 51.1 New Zealand 19.2

Cyprus 49.3 Japan 18.9

Great Britain 48.3 Israel 18.7

Germany 47.5 Switzerland 18.2

Japan 47 Iceland 17.6

Portugal 39.9 Finland 17.2

Israel 39.4 Sweden 14.8

South Korea 27.2 Cyprus 13.6

France 25.9 Norway 10.6

Conclusion

Precipitous declines in trust and faith in opportunities and

meritocratic rewards are the biggest threat to social

cohesion in the UK.

Education can play a major role in equalising

opportunities and counteracting the erosion of core beliefs

which hold society together. But at the moment it is not

doing this .

Policy needs to concern itself not only with raising average

levels of skills but equally with how lifelong learning systems

spread skills around.

References

Green, Preston and Janmaat (2006) ‘Education,

Equality and Social Cohesion’, Palgrave.

Green and Janmaat (2011) ‘Regimes of Social

Cohesion: Societies and the Crisis of

Globalisation’, Palgrave.

Llakes.org