Whether referenda should be adopted to decide questions about national-identity/boundary issues has been debated for the last two centuries. The German lawyers Hotzendorf, Geffker, Stoerk, and Francis Liever, for example, denied use of referenda on the grounds that they subjected the minority to the rule of a simple majority without protection; that they contradicted the organic nature of the state; that they would encourage secession and make the establishment of peace more difficult; and that they were used to ratify faits accomplis. (1)
Likewise, British ministers and their constitutional advisers were reluctant to use referenda to smooth their own path toward decolonization because (1) referenda would undermine representative parliamentary systems; (2) there was a perceived lack of "readiness" among the people concerned; and (3) there were the difficulties of providing fair protection for various ethnic groups living in the same country. The British institutional arrangements were limited franchises and balanced constituencies. (2)
Kedourie further points out the difference between elections of leadership and plebiscites (or referenda) over national-boundary questions. Plebiscites and referenda express a decision made once and for all, yet their results are no less conclusive or arbitrary than those of any other methods. For Kedourie,
Plebiscites are not more certain, more equitable, or less liable to criticism than the traditional methods by which boundaries are determined and which are based on the balance of power and the compromise of conflicting interests. (3)
Kedourie dismissed Renen's idea of a daily plebiscite as inadequate, "for a political community which conducts daily plebiscites must soon fall into querulous anarchy, or hypnotic obedience." (4) Cobban also presents a different kind of argument against plebiscites: in the history of referendum, the employment of plebiscites are contingent, contextual, and uncertain. (5)
However, referenda as a procedure, or method, appears preferable to methods such as conquest by force or arbitrary decisionmaking by a group of powerful elites. A referendum is the only satisfactory method by which the will of the people can be ascertained. In 1933, Wambaugh offered a convincing argument in this regard:
There is, however, no perfect method of establishing national boundaries. The problem is one of alternatives, a choice between methods varying in imperfection. To allow questions of sovereignty to be settled by conquest, or by a group of Great Powers gathered at a Peace Conference, resorting for their method of determination at one time to strategic considerations, at another to languages statistics, or to history, or to geographic or economic criteria--such methods are even less satisfactory to democratic principles. Therefore it seems certain that we shall keep the plebiscite as a tool in the workshop of political science. (6)
Recently, Beran developed a democratic theory of secession, seeing referenda over identity/boundary questions as part of the democratic management of the identity/boundary question, an important mechanism of voluntary agreement and consent. He also imagines an ideal world order in which referenda can settle any dispute over boundary questions, in particular the question of secession. (7)
This article attempts to investigate empirically to what extent reality achieves, or falls short of, this ideal. Through an examination of all 173 boundary-related referenda in the period 1791 to 1998, the article aims to test and further develop the democratic theories of the national-identity/boundary question and to identify a set of conditions under which a referendum can be held successfully to decide the boundary issue. The national-identity/boundary problem is inextricably bound up with such issues as autonomy, secession, independence, and reunification. Therefore, this article examines only referenda related to, or used to decide, the identity/ boundary questions that deal with those issues.
It is acknowledged from the outset that referenda are not the perfect solution to all boundary problems and that there is no perfect method of settling the national-boundary question. Moreover, only free and fair referenda are democratic because the settlement or resolution of any national-identity/boundary question is based on the consent of the people, whose votes ultimately decide the outcome of the boundary question. Some referenda are undemocratic, however, because they are unfair and coercive and are manipulated by politicians to provide legitimacy for leaders' claims over territories. Such referenda cannot help to resolve the national-identity/boundary question.
To address the question of whether referenda actually decide the outcome of the boundary question, it is also necessary to qualify the term decide. Decide here refers to a decisive mechanism whereby the people cast their votes on a specific boundary issue: the outcome of the referendum is so binding that it is capable of settling the boundary question. For example, Wales failed to gain autonomy in its 1979 referendum because only 20.9 percent of voters voted in favor of it; however, in a 1997 referendum, autonomy was supported by 50.3 percent of the voters and a Welsh assembly was successfully established.
In other cases, the outcomes of referenda are not always binding and decisive. The French government, for example, rejected the results of a 1919 Luxembourg referendum in which 71 percent of voters were in favor of establishing a customs union with France rather than Belgium. Another form is the nonbinding or unofficial referendum, such as the unofficial 1921 vote in Tyrol over the question of whether it should unify with Germany and the unofficial 1919 vote in the Aaland Isles over the question of whether they should be unified with Sweden. In other cases, referenda serve only to rubberstamp a solution to the boundary question that has already been decided either by political forces or by parliamentary vote/declaration.
Historical Overview of Boundary-Related Referenda
The referendum, as a principle and procedure, was invented during the French Revolution and came to be largely accepted in the democratic West. Europe, with its well-established democracies, has the highest number of boundary-related referenda occurring during the period between 1791 and 1998. Nearly half of the referenda over the boundary question have been held in Europe. (8) Theoretically, the idea of democracy presupposes the idea of popular sovereignty, which in turn supports the project of democratic management of the boundary question.
France was the first modern state to exercise democratic, principles to resolve the national-boundary question. (9) The union of Avignon, Savoy, and Nice with France was accomplished through referenda. The referendum principle was rejected, however, by the bayonets of Napoleon. It was again rejected on the Alsace-Lorraine issue in 1919 after France and its allies won World War I. However, in the 1950s and 1960s, the democratic Fifth Republic of France used referenda widely to settle boundary questions in the process of decolonization.
The United States, on the other hand, did not bother with referenda when annexing Florida, Louisiana, California, Texas, and other territories; and the Civil War thwarted the bid for self-determination by the Southern states. However, the United States later supported the concept of self-determination. (10) President Wilson was a particularly strong advocate and promoter of the self-determination principle and democratic management of the boundary question. In 1916, he said that all peoples have the right to choose the sovereignty under which they should live. (11) The Marshall Islands, the Federated States of Micronesia, and Palau all gained independence from the United States through referenda.
In 1859, the British government supported the idea of an Italian referendum on the unification of the provinces of Sardinia and Lombardy with the provinces of Tuscany and Emilia. At the 1864 London Conference, the British government urged the use of a plebiscite to convert into lasting peace the temporary truce between Denmark and the Germanic powers over Schleswig-Holstein; (12) and after World War I, the British government proposed plebiscites in Allenstein on the German-Polish border and in upper Silesia. (13) However, British ministers and their constitutional advisers were reluctant to use referenda to smooth their own path toward decolonization because, firstly, it was felt that referenda were likely to undermine the representative parliamentary system; secondly, there was a perceived lack of "readiness" of the people concerned; and thirdly, there were the difficulties of providing fair protection for ethnic groups when various groups lived in the same country. (14) Antigua, Barbuda, and Saint Kitts-Ne vis gained independence from the United Kingdom through parliamentary vote; Belize, Dominica, Grenada, Saint Lucia, Saint Vincent and the Grenadines, the Seychelles, and the Solomon Islands achieved the same result through negotiated settlement. It should be pointed out that the reluctance to use referenda does not imply that the British government rejected democratic management, but that they preferred the parliamentary mode of democracy. As already pointed out, the British government has employed referenda to help decide the boundary question in Northern Ireland (in 1973 and 1998), to settle the autonomy question over separate parliaments or assemblies in Scotland and Wales (in 1997), and to settle the issue of whether the United Kingdom should join the European Union (in 1975).
The 1848 unification of Italy resulted from referenda, but Italy denied referenda in the wake of World War I because it doubted whether the people of the Tyrol, south of the Brenner Pass, and in large parts of the Dalmatian coast (areas secured through secret...