Horizontal Inequality: Two Traps

1
Horizontal Inequality Two Traps

• By Frances Stewart

2
Many issues to do with defining inequality

• Inequality among whom?
• Inequality of what?
• Inequality over what period?
• Here concerned with persistent inequality among
  groups in multiple dimensions in capability
  space.
• I shall identify two types of inequality trap
  that underlay such persistent group inequalities.

3
Issues to be covered

• Definitional issues
• What HIs are
• Multidimensionality
• Why HIs matter
• How persistence is defined.
• Evidence of persistence in HIs
• Reasons for such persistence two inequality
  traps.
• Evidence for these traps.
• Some policy implications.

4
Horizontal Inequality defined

• HI is inequality between groups as against
  between individuals or households (vertical
  inequality).
• Which groups? groups with meaning to members,
  viewed by people themselves, or others as
  important aspect of identity.
• Group boundaries socially constructed and vary
  across societies and over time.
• Examples of salient identities
• Ethnic/tribe African
• Religious most regions notable N.Ireland
  Middle East
• Race e.g. South Africa Malaysia Fiji
• Regional (overlaps with other identities)
• Caste (South Asia)
• Class

5
Multidimensionality

• Major dimensions socio-economic political
  cultural status.
• Sen answered question, inequality of what in
  terms of capabilities -- being and doing valuable
  things.
• Here too concerned with inequalities in multiple
  capabilities

6
Why do HIs matter?

• Direct impact on Wellbeing.
• Group inequality affects peoples well-being.
  Being Black and feeling Blue.
• poverty is bad enough, but when you are being
  discriminated, this strips away your dignity, it
  is much worse, you feel humiliated, you feel
  useless (former member of Guatemalan guerrilla
  group).
• Embodied in Akerlof and Kranton welfare
  function.
• Instrumental.
• Affects economic growth. Group handicap prevents
  efficient solutions. Unequal access to assets,
  markets, social capital.
• Affects poverty. If discrimination major reason
  for poverty, group-blind policies may not work.
• Political Instability evidence that HIs
  increase risk of violent conflict

7
Political instability and HIs

• Evidence in many countries N.Ireland Nepal
  Chiapas Mexico Sudan
• Econometric cross-country evidence (Ostby
  Barrows) and within country (Mancini Murshed
  and Gates)
• Note groups readily mobilised if suffer
  socio-economic HIs.
• Leaders mobilise if suffer political HIs.
• See Horizontal Inequalities and Conflict
  Understanding Group Violence in Multiethnic
  Societies (F.Stewart ed).

8
Persistence defined

• Persistent group inequality can occur over many
  generations.
• Persistent group inequality has great negative
  impact on deprived because it limits prospective
  upward mobility.
• Relevant time span for persistent HIs is period
  over which group boundaries remain salient.
  Sometimes boundaries (or salience of them)
  disappear and sometimes inequalities disappear.

9
Evidence of persistence of HIs many examples

• N. Ireland Catholic/Protestant inequalities back
  to 17th century.
• India Muslim/Hindu throughout 20th C.
• US black/white since 17th century.
• Northern Ghana IMR higher than South since
  1930s.
• South Africa black/white inequalities back to
  white migration.
• But note we notice persistent HIs non-persistent
  HIs also (e.g. many immigrant groups).

10
Evidence of persistence of HIs also from strong
tendency for intergenerational transmission of
incomes and education

• Transmission irrespective of group (most
  research). Strong correlation between parents and
  childrens education (and also incomes). Over 42
  countries, 1 standard deviation difference in
  parental education associated with 0.4 of
  subsequent generations education.
• Correlation of 0.6 parents and children LA
• Asian and Western countries, 0.4.
• UK, US, half variation in incomes explained by
  parental income. Lower in Scandinavia.

11
Intergenerational transmission by group fewer
studies

• US 42 of blacks born at bottom decile remain
  there only 17 of whites.
• Brazil younger blacks do better than older
  blacks, but no better relative to white age
  cohort.
• South Africa poorest blacks have lowest
  mobility, and mobility greater for whites than
  blacks.
• India education of both Ms and Hs strongly
  related to parents, but Ms show downward mobility
  e.g. 71 of Ms have less education than a
  parent with full secondary education only 12 of
  Hs.
• In summary tendency for persistence in income
  and educ. rank irrespective of group, but
  stronger immobility among members of deprived
  groups.

12
Why? Two traps Trap 1 the capability trap

• Interactions among capabilities mean that lack of
  one capability makes it difficult to escape from
  a low position and easy to maintain a high one
  e.g. poor education leads to poor health leads to
  poor income, leads to poor education and health

13
Arrows show how one type of capability affects
others and how each affect and are affected by
incomes
14
Trap 2 the capital trap

• Interactions among types of capital similarly
  cause an inequality trap productivity of one
  type of capital depends on access to others e.g.
  financial capital or physical capital better used
  with human capital human capital better used
  with financial capital and all more productive
  with good social and cultural capital.

15
Arrows show how access to each type of capital
affects returns to others
16
Why does group inequality persist even more than
individual poverty in homogeneous societies?

• Social and cultural capital key aspects of
  persistent group inequality
• Asymmetric social capital, i.e. contacts and
  networks, a critical aspect of groups (more
  intra-group, less intergroup contacts Blau).
• Cultural capital (Bourdieu) concerns aspirations
  and behaviour. Also tends to differ between
  groups.
• Plus discrimination (a form of negative
  social/cultural capital)
• Consequently even if reduce inequality in other
  capitals, persistent productivity differences
  likely to remain.
• And political inequalities can cause or reinforce
  differences in all types of capital.

17
Evidence of traps1. Capability interrelationships

• Much empirical evidence that
• Female education improving childrens health and
  nutrition, including by reducing fertility
• Education increases childrens education.
• Improved health and nutrition improves
  educational outcomes.
• Improved health and nutrition positive impact on
  productivity and earnings.
• More education raises earnings.
• More earnings contribute to improved education
  and health and nutrition.
• Evidence from Schultz, Thomas, Behrman, Wolfe,
  Chaudury, Immink and Viterri, haddad and Bpuis,
  Psacharapolous, Glewwe, Bloom, Alderman,
  Strauss.

18
2. Evidence on interrelationships among capitals

• Deprived groups generally have less of several
  types of capital - examples
• Indian Muslims compared with Hindus have less
  human capital, bank credit, poorer
  infrastructure, poorer social contacts, place
  less value on education (Bhalotra, et al).
• Ethnic minorities in Vietnam have less schooling,
  less access to credit, fewer farm tools, and
  worse quality land than others (Baulch et al).
• US blacks have less education, poorer health,
  fewer financial assets and worse infrastructure

19
Evidence on asymmetric social capital
Source CRISE Perception Surveys.
20
Returns to education worse for deprived groups

• Evidence from Peru, Vietnam, Brazil, Uttar
  Pradesh.
• Attributed to poorer quality schools (public
  capital), poorer support and attitudes from
  households (weaker cultural capital), poorer
  contacts for choosing schools and jobs (poorer
  social capital).

21
Figueroas illustration Peru. Differential
returns to education
Returns to education
White population
Mestisto population
Indigenous population
Years of education
22
Implications of twin traps for policy

• Multiple deprivations mean that most individual
  policy interventions unlikely to be successful.
• Market alone inadequate because of multiple
  handicaps.
• State action can (try to) eliminate current
  discrimination, but past discrimination handicaps
  current generation.
• Action on one capability e.g. education
  insufficient because of deficiencies in others.
• Coordinated action on health, education etc.
  better, but capital trap remains.
• Taking action on access to finance, land etc.
  helpful, but asymmetric social and cultural
  capital persists.
• Problems clear from persistence in HIs even in
  societies where some action is being taken, e.g.
  US, Australia, Canada, Brazil, South Africa.

23
Effective policy needs to address multiple
deprivations in capabilities and capitals

• This implies
• Eliminating current discrimination.
• Eliminating inequalities in access to public
  services
• Redistributing financial and land assets.
• Overcoming social and cultural capital
  asymmetries by positive efforts to ensure access
  to good schools, good jobs etc.

24
Is this possible?Positive examples

• Malaysia across the board action on education,
  employment and assets. Catch up but after nearly
  40 years, big gaps remain (average income Malays
  60 of Chinese share ownership 2).
• N.Ireland action on education, employment and
  public services. Largely successful.
• Elsewhere partial action has achieved much less
  e,g, Ghana, Nigeria, US, Canada, Australia, New
  Zealand.

25
Conclusion

• First need to acknowledge importance of issue.
• Monitor outcomes and particular deprivations.
• Often issue is ignored.
• Political constraints remain critical.