Excerpt from Introduction
"Scholars in development studies increasingly recognize that inequalities between groups constitute a more potent source for
violent conflict than inequalities among individuals (Stewart, 2000; Chua, 2002; Justino, 2003; Ndikumana, 2004). When
inequalities in incomes, wealth, and access to social services or political power coincide with group differences, ethnicity
may assume importance in shaping choices and mobilizing individuals for collective action. Yet little is known about the
nature and dynamics of ethnic cleavages and inequalities, especially as they affect the public
sector, which plays a central role in resource allocation and identity formation. The
public sector may be rendered ineffective or illegitimate if it fails to develop
mechanisms to regulate difference and ensure inclusiveness.
This paper examines the complex ways ethnic cleavages and inequalities affect public
sector institutions. It focuses on issues of inclusion, accommodation and cohesion in
the constitution and management of the public sectors of multiethnic societies under
formal democratic rule. Ethnicity affects the identities of states, access to institutions,
and the confidence different sections of society may have in government. Inequalities
arising from the structure of opportunities and the way the public sector is governed
are often a source of tension as individuals use group solidarity to maintain or alter
advantages.
The paper challenges three influential ideas about ethnicity, governance and cohesion.
First, it is often believed by development analysts, political theorists, policy makers
and the lay public that countries that are ethnically diverse are likely to be less
cohesive and stable (Barry, 1991; Connor, 1994; and Miller, 1995), more prone to
violent conflict (Furnivall, 1948; Smith, 1969; Ignatieff, 1993; and Moyniham, 1993),
less disposed towards democratic government (Miller, 1995; Barry 1991; Nyerere,
1967; Museveni, 1997), more likely to experience low levels of industrialisation
(Gellner, 1983) or growth (Easterly and Levine, 1997), and less likely to sustain
universal welfare programmes (Goodhart, 2004). Although some recent quantitative
studies (Collier and Hoeffler, 2001a,b; Elbadawi and Sambanis, 2001; Fearon and
Laitin, 2003; Mozaffar, Scarritt and Galaich, 2003) have challenged some of these
assumptions, the alternative views they posit have tended to minimise the role of
diversity in influencing social and political outcomes. We argue that the relevant issue
is not the existence of diversity per se, but types of diversity, which can constrain or
support particular outcomes. Ethnic fragmentation does not necessarily produce
pathological situations; nor is it helpful to downplay diversity in explaining social
behaviour. Ethnic cleavages are configured differently in different social structures
and are less conflictual in some countries than in others.
Second, contrary to the assumptions of liberal individualism, it is difficult to achieve
proportionality or balance in the public sector if policies are not oriented towards it.
Multiethnic societies that adopt ethnicity-blind policies run the risk of creating highly
unequal public sectors. Even a cleavage-neutral policy, such as merit-based
recruitment or one based on republican civic values, may produce unequal outcomes
because of the unequal starting points of groups. Our study suggests that relative
balance has been achieved in countries that are highly fragmented or those with
ethnicity-sensitive policies that are oriented towards high levels of proportionality.
Some countries have achieved less disproportionality in some institutions because of
redistributive policies that focus on those institutions. Most of our cases display
varying levels of inequality and weakly structured or non-existent policies for
promoting balance.
Third, politicians and citizens face different types of constraints in constituting the
public sector. These stem from the dynamic interplay of the structures of ethnicity,
group dynamics and inequalities. Institutions, however well crafted, may have
different levels of significance in different social settings. Ethnic structures with two
or three key groups often require ethnically sensitive institutions that provide
incentives to cooperate and avoid conflicts. Ethnicity-sensitive institutions may not be
relevant in situations where there is one overwhelmingly large group or where the
ethnic structure is highly fragmented. In this regard, the paper challenges two policy
frameworks that have been held up as solutions to the governance problems of
ethnically divided societies. These are majoritarian policies that reward moderation in
party behaviour and vote pooling while also encouraging adversarial politics; and
consensus-based or power sharing arrangements that seek to accommodate the ethnic
segments. The first seeks to promote plurality within the party system by encouraging
actors to seek votes outside their ethnic strongholds (Horowitz, 1990). The second
accepts ethnic-based parties as given, and promotes plurality at the governmental
level rather than in the party system (Lijphart, 1999). Our study suggests that although
the pulls of majoritarian rule and power sharing are very strong, they do not always
pull in opposite directions. The majority of ethnically segmented countries, like their
more homogenous counterparts, have opted for majoritarian solutions. However,
ethnic problems have forced some of them to incorporate power-sharing elements in
their majoritarian institutions.
The next section of the paper discusses the typology that guides the study. Section
three provides a conceptual discussion of public sector institutions as they relate to
issues of identities and access. Section four, which is divide into five parts, analyses
ethnic cleavages, group dynamics and inequalities in four public sector institutions:
cabinet, parliament, civil service and party system. Section five examines institutions
for managing diversity and inequality. The institutions range from electoral rules to
governance arrangements for power sharing, federalism and decentralisation, and
protection of minority rights. Redistributive policies and rules that seek to correct
disproportionality are also analysed. An index of ethnic proportionality is developed
to guide the discussion on the relative inclusiveness of the different public sectors. Fifteen
countries that formed the core of the UNRISD project on Ethnic Structure, Inequality
and Governance of the Public Sector provide the basis for the analysis. These countries are Belgium, Bosnia-Herzegovina, Botswana, Ghana, Fiji, India, Kenya, Latvia, Lithuania, Malaysia, Nigeria, Papua New Guinea, Switzerland, Tanzania and Trinidad and Tobago."
