21st September 1997


Home Page Front Page OP/ED News Business

Devolution: rights and responsibilities

By Walter Ladduwahetty

Robert C. Oberst in an article published in the Daily News of March 14th 1997, commenting on the ongoing devolution debate states that “Many of these writers do not seem to understand that democracy is not simply majority rule... Every successful democracy in multi-ethnic societies has been forced to create mechanisms in their political system to prevent simple majorities from ruling... When simple majorities rule, some minorities may for ever lose the right to be represented by “their” government. This is true whether the minority is ethnic in nature or whether it is social or class based...”.

Similar views were expressed by Sumanasiri Liyanage in an article published in the Daily News of December 17th, 1996. He stated that “... in pluralist societies like India, Malaysia and Sri Lanka, democracy in its primary meaning cannot be established through a system ensuring majoritarian rule. What is necessary is to create political institutions and to re-structure the state ensuring that those who are affected by a decision get a fair chance to participate in the decision making process ... In the last fifty years, the Sri Lankan state has gone through a process of Sinhalization which resulted in the political exclusion of Tamils in the legislative and executive process. So the so-called ethnic question is none other than a problem emanating from the political exclusion of one group comprising the population “.

According to Oberst, “The most effective methods of protecting the minorities has taken two forms, federalism or devolution, and consociationalism. Consociationalism is ... best suited to European cultures. Federalism and devolution have been applied in Europe and the non-European world with a great deal of success”. Commenting on the current devolution proposals, Liyanage states “Since independence this is the first major step to make a constitutional framework to accommodate the interests of the minority communities. This marks a deviation from the notion of majoritarian democracy which up to now has determined the democratic set-up in Sri Lanka”.

“Federalism or devolution” may have served as an effective anti-majoritarian mechanisms in another time and another place. But, wherever it served effectively, the federal regions represented near contiguous homogeneous groups. The demographics in Sri Lanka does not lend itself to the creation of contiguous homogeneous federal units. The federal units proposed instead, will in effect create two majorities, each burdened with the complexities of ethnic issues. The question as to how the Sinhalese majority and the Tamil majority regions will deal with their respective minorities is not addressed and remains unresolved. Therefore, although the intent of the solution was to find an antimajoritarian mechanism, “federalism or devolution” leaves the minorities helpless in the hands of two majorities.


If as stated, majoritarian rule is the primary cause for the conflict in Sri Lanka, are there workable safeguards or alternatives that could be implemented? According to some, “... the problem posed by majority rule and the alternatives to it is one of extreme difficulty for which no completely satisfactory solution has yet been found” (Dahl 1989, “Democracy and its Critics”). Continuing, he adds “... we are entitled to be just as skeptical about claims that an alternative would be clearly superior to majority rule or more consistent with the democratic process and its values. For all the alternatives to majority rule are also seriously flawed” (Ibid).

Nordlinger identified “six Conflict-Regulating Practices”, but did not include FEDERALISM as one of the “Practices”. The reason for doing so was because “... compared to the six conflict-regulating practices, federalism may actually contribute to a conflict’s exacerbation and the failure of conflict regulation. In some deeply divided societies it is impossible to draw state boundaries without including a large number of individuals belonging to segments whose territorial base is elsewhere. Federalism thus allows or encourages the dominant segment in one state to ignore or negate the demands of the minority segment. The possible consequences vary, but they may very well lead to the conflict’s further exacerbation, as happened with an Ibo minority living in Northern Nigeria. Moreover the combination of territorially distinctive segments and federalism’s grant of partial autonomy sometimes provides additional impetus to demands for greater autonomy; when the centrally-situated or centrally-oriented conflict group refuses these demands, secession and civil war may follow” (Nordlinger 10972", Conflict Regulation in Divided Societies).

Commenting on Nordlinger’s observations that federalism could lead to secession or partition, Lijphart ALSO does not rule out such possibilities, and in fact states that “secession should not be regarded as an undesirable result of the tensions in a plural society under all circumstances” especially if assimilation or consociational solutions have been tried and failed. “...the remaining logical alternative is to reduce the pluralism by dividing the state into two or more separate and more homogeneous states” (Lijphart l977, “Democracy in Plural Societies”).

As an alternative to majoritarian rule, Lijphart proposes the consensus model of democracy which he “describes in terms of eight elements”, ONLY ONE of which is “Territorial and non-territorial federalism and decentralization” (Lijphart 1984, “Democracies”). In other words, Lijphart treats territorial and non-territorial federalism within the framework of consociationalism and not as an alternative to it, because the exercise of constitutional privilege by one or more federal units could have adverse effects on a majority of the other units, in which event the exercise of constitutional privilege would NOT be consensual and therefore undemocratic. This aspect is clearly demonstrated by Dhal: “In a federal system, in some cases the minority would prevail, and the national majority could do nothing about it, constitutionally speaking.... In these circumstances, a unitary system - or an anemic federal one - satisfies the criteria of the democratic process better than a robust federal system” (Dhal 1989, “Democracy and its Critics”).

The lesson from all the opinions cited is that alternatives to majoritarian rule contain their own brand of short comings that can prove to be undemocratic and unpalatable to a section of the people regardless of whether the state is unitary or federal. Even a minority safeguard such as the right to veto if exercised indiscriminately can block even a supramajority from enjoying certain rights and therefore prove to be an undemocratic device. Similarly, majoritarian rule if exercised without thought and concern for those who would be affected by a decision is morally wrong even though it may be their democratic right.

Devolution for Sri Lanka

The Sri Lankan society is made up of 74% Sinhalese, 12.6% Sri Lankan Tamils, 7.1% Muslims, 5.6% Indian Tamils and 0.7% others (1981 Census). If the Northern and Eastern provinces are merged to create the territorial base desired by the Tamil community, the Sri Lankan Tamils would be 65.1%, the Sinhalese would be 13.2%, and the Muslims would be l7.6 , with the Indian Tamils being 3.6%, and the others being 0.4%. As far as the remaining seven provinces are concerned the Sinhalese would constitute 83.9%, and the rest collectively would be 16.1%.

The ethnic composition of the Sri Lankan Tamils in the merged North and East would be 65% of the population, and therefore a majority, but not to the extent of the Sinhalese majority of 83.9% in the remaining provinces. The devolution proposal would amount to the creation of one merged region with a Tamil majority, and another region where the Sinhalese by most counts could be considered homogeneous, or at least a supramajority regarding issues relating to ethnicity. Creating two majorities leaves the minorities of each region exposed to the rule of two majorities without the protection of anti-majoritarian mechanisms in either region.

Despite the Sri Lankan Tamils being 65% in the merged north and the east, the ground situation is such that they are 75% of the population in only three of a total of seven districts. In two of the remaining four districts, the Sri Lankan Tamils are in fact an ethnic minority and in the rest they enjoy a marginal majority. In addition, only two of the districts with sizable majorities are contiguous (Jaffna and Mullaitivu) and the third district (Batticaloa) is isolated and separated by a district where they are a minority. As a result, the Sri Lankan Tamil ethnic concentrations are not contiguous. By contrast, the Sinhalese are contiguous and a sizeable majority (83.9%) in all the remaining districts except one where no ethnic group enjoys a majority.

The Sri Lankan Tamils of the eastern province do not form a simple majority on their own. The only method by which they can be elevated to a majority status is to couple them with the population of the north. The outcome of this coupling is to merge the two provinces, and in the process to annex to themselves 50% of the land in the province that is currently occupied by the Sinhalese. The statement in the Indo-Sri Lanka agreement “that the Northern and Eastern Province have been areas of historical habitation of Sri Lankan Tamil speaking peoples...” ignores the fact that they have been “historically “ confined to the coastal belt as they are even to this very day. All the remaining land in the eastern province (50%), is what is left to ancestral land (“purana” villages) that at one time belonged to the indigenous Sinhalese people. Annexation of this land by a merger is a violation of the rights of these indigenous people.

The eastern province taken as a whole is unsuitable for territorial federalism because of the absence of an ethnic majority. Furthermore, because the ethnic distributions are interspersed over the province, near homogeneous ethnic regions cannot be created. The pattern that exists is one of pockets of high ethnic concentration within each AGA division, sometimes clustered together and at times completely isolated. Ignoring this peculiar pattern of ethnic composition and expressing ethnic compositions in terms of generalities of percentages of population in a merged north and east can be deceptive, and could mislead one to come to the conclusion of the suitability for territorial federalism.

The ground situation in regard to ethnic distribution in Sri Lanka is unlike in other countries, and therefore does NOT lend itself to “territorial” or “nonterritorial federalism” (Lijphart 1984). This is particularly so of the Muslim community in Sri Lanka, where 30% are in the eastern province of which half are concentrated in six AGA divisions, five of which are contiguous. The remaining 70% occupy pockets of concentration throughout the island. It is because of the unique nature of ethnic concentration that the Muslim community of the eastern province objects to be associated with the territorial unit as curently proposed. This group of Muslims is in fact seeking a separate arrangement for themselves.

Devolution or straightforward federalism as an anti-majoritarian mechanism is too simplistic an approach, and unworkable in the Sri Lankan context. “And where federalism has been employed, as in Nigeria, Mali, East Africa, Ethiopia, and the Congo Republic, the results have not been notable for their enduring qualities. Federal systems have remained operative for relatively brief periods of time, followed by fissure into separate, sovereign parts or movements towards unitary systems. Federalism has proved brittle” (Nordlinger 1972, quoing Rothschild 1966, “The Limits of Federalism”).


The devolution proposals aim at creating a unit of regional power where the Sri Lankan Tamils will be a majority because of their ethnic concentration in the merged provinces of the north and east. The impression that an ethnic concentration exists is arrived at statistically by evenly spreading the 60% of the Sri Lankan Tamils who were concentrated in the district of Jaffna (1981 Census) over the entirety of the merged north and east. As a result the presumed impression of ethnic concentration is illusionary. Furthermore, the conclusion ignores the unique demographics in the eastern province. As illustrated already, the demographics are such that ethnic cleavages do not relate to the territorial boundaries. In this context, a territorial approach is unsuitable and therefore cannot be justified.

A durable solution to the conflict in Sri Lanka must necessarily be inspired by principles and values of democracy. Borrowing Liyanage’s quotation of W. Arthur Lewis, “the primary meaning of democracy is that all who are affected by a decision should have the chance to patricipate in making that decision...”. This “primary meaning” demands that circumstances are created and the opportunity given, for the “demos” of the eastern province to freely express their consent as to whether they want to be associated within a political framework of a merged north and east province, or whether they want to be separate and distinct from the north. To deny the opportunity to participate in such a momentous decision because of Tamil insistence would be a mockery of the “primary meaning” of democracy.

The provision to hold a referendum “... to enable the people of the Eastern Province to decide... “ is embodied in the Indo-Sri Lanka agreement of 1987. When the Sri Lankan Tamil community insists that the two provinces be merged, not only is a democratic right guaranteed jointly with another sovereign power being violated, it is also a violation of the terms of an international agreement that could have serious repercussions. Yielding to the demands of the Sri Lankan Tamil community’s insistence to merge the two provinces would be to submit to the tyranny of a minority.

If the outcome of a referendum is not to merge the two provinces, there is a possibility that the Sri Lankan Tamil community may not be interested in devolution as a solution. The resolution to the conflict then may take a different track. Instead of a territorial approach, which is one of exclusion, approaches that are more inclusive such as sharing power at the centre may become a realistic and attainable possibility. Such an approach could take the form of a committee system for the legislature where all the elected members of the legislature are involved in the governing process, and a nominated cabinet of ministers who could represent the ethnic ratio of the country.

It appears that meaningful alternatives to majoritarian rule ultimately depend on the commitment of both the majorities and minorities: on both realizing that their well-being, their future, and their security are inextricably linked together and not separately. This can only happen if the political leadership of both groups realize that just as much as each has rights and entitlements, each also has responsibilities to each other. It is a balance of these two that lead to stability.

Continue to Plus page 7

Return to the Plus contents page

Read Letters to the Editor

Go to the Plus Archive



Please send your comments and suggestions on this web site to or to