How Independence Referendums Work

A referendum is the most direct way a people can decide its political status. Every vote on independence also answers a set of practical questions before the first ballot is cast. Who writes the question? What margin counts as a decision? Who counts the votes, and who watches the count? What happens the morning after?

The official records of five votes in four places give concrete answers.

Scotland, 2014

The vote began with an agreement between two governments. On October 15, 2012, the UK Government and the Scottish Government signed what the Electoral Commission calls an agreement "to work together to ensure that a referendum on independence for Scotland could take place."

An order under the Scotland Act then gave the Scottish Parliament the power to hold the vote, on conditions: the poll had to be held "no later than 31st December 2014," and "the ballot paper must give the voter a choice between only two responses."

The Scottish Government proposed this question: "Do you agree that Scotland should be an independent country?" The Electoral Commission, which the two governments had charged with advising on the question, tested it with voters. In January 2013 it found the question easy to understand but "not neutral because the phrase 'Do you agree ...?' could lead people towards voting 'yes'." It recommended a shorter form, and the Scottish Parliament adopted it word for word. Section 1 of the Scottish Independence Referendum Act 2013 reads: "Should Scotland be an independent country?"

The Act set no minimum turnout and no supermajority. The poll was held on September 18, 2014. According to the Electoral Management Board for Scotland, 4,283,938 people were eligible and 3,623,344 voted, a turnout of 84.6 percent. The result:

Option	Votes	Share
Yes	1,617,989	44.7%
No	2,001,926	55.3%

Scotland stayed in the United Kingdom. The question did not go away. In 2022 the Scottish Government's senior law officer asked the UK Supreme Court whether the Scottish Parliament could hold a second vote on its own authority. The Court noted that in 2013 an Order in Council had enabled the 2014 vote and that the UK Government was "currently unwilling to agree to the making of another Order in Council." It ruled unanimously that, without such a change, "the Scottish Parliament does not have the power to legislate for a referendum on Scottish independence."

Quebec, 1980 and 1995

In Quebec the provincial legislature set the question. Under the Referendum Act of 1978, "On a motion of the Prime Minister, the National Assembly may adopt the text of a question which is to be the subject of a referendum." Élections Québec, which publishes the official results of both votes, notes that "In Québec, referendums are consultative."

The 1980 question was long. It asked voters for a mandate to negotiate, and it promised a second vote before any change:

"The Government of Québec has made public its proposal to negotiate a new agreement with the rest of Canada, based on the equality of nations; this agreement would enable Québec to acquire the exclusive power to make its laws, levy its taxes and establish relations abroad – in other words, sovereignty – and at the same time, to maintain with Canada an economic association including a common currency; no change in political status resulting from these negotiations will be effected without approval by the people through another referendum; on these terms, do you give the Government of Québec the mandate to negotiate the proposed agreement between Québec and Canada?"

On May 20, 1980, 3,738,854 of 4,367,584 registered electors voted, a turnout of 85.61 percent. Yes received 1,485,851 votes and No received 2,187,991. No won 59.56 percent of valid ballots.

The 1995 question was shorter but still tied to other documents:

"Do you agree that Québec should become sovereign, after having formally offered Canada a new economic and political partnership under the bill respecting the future of Québec and the agreement signed on June 12, 1995?"

On October 30, 1995, turnout reached 93.52 percent of 5,087,009 registered electors. Yes received 2,308,360 votes and No received 2,362,648. No won 50.58 percent of valid ballots, a margin of 54,288 votes.

Two answers about process followed the 1995 vote. In 1998 the Supreme Court of Canada held that a vote, to count as an expression of the democratic will, "must be free of ambiguity both in terms of the question asked and in terms of the support it achieves." It held that a clear expression of that desire "would give rise to a reciprocal obligation on all parties to Confederation to negotiate constitutional changes to respond to that desire." It left the exact standard to elected officials: "it will be for the political actors to determine what constitutes 'a clear majority on a clear question.'"

In 2000 the Parliament of Canada passed the Clarity Act. It directs the House of Commons to judge the clarity of any such question. It states that a clear result could not come from a question "that merely focuses on a mandate to negotiate," or from one that "envisages other possibilities" such as economic or political arrangements with Canada. After a vote, the House must weigh "the size of the majority of valid votes cast" and "the percentage of eligible voters voting in the referendum."

Přemysl Rosůlek's study in this library, Secession and Intelligibility of a Referendum Question, compares how the Quebec and Scottish questions were worded and how clearly voters could understand them.

The United Kingdom and the European Union, 2016

The 2016 vote concerned a sovereign state leaving a union of states. Its rules still show how a people can decide its status.

Parliament wrote the question into law. Section 1 of the European Union Referendum Act 2015 sets it out: "Should the United Kingdom remain a member of the European Union or leave the European Union?" The two answers on the ballot were "Remain a member of the European Union" and "Leave the European Union." The Act set no minimum turnout and no supermajority.

The vote was held on June 23, 2016. The House of Commons Library reports these totals:

Option	Votes	Share of valid votes
Leave	17,410,742	51.9%
Remain	16,141,241	48.1%

A total of 33,577,342 people voted, a turnout of 72.2 percent. Leave won by 1,269,501 votes. The result differed by country. Leave won majorities in England and Wales. Remain won 62.0 percent in Scotland and 55.8 percent in Northern Ireland.

The United Kingdom left. The European Union (Withdrawal) Act 2018 fixed "exit day" as "31 January 2020 at 11.00 p.m." The departure came three years and seven months after the vote.

Montenegro, 2006

Montenegro's vote ended in independence. Its rules were negotiated in detail before the campaign began.

Montenegro was then part of the State Union of Serbia and Montenegro. According to the observation mission of the OSCE's Office for Democratic Institutions and Human Rights (ODIHR), the 2002 agreement that created the State Union placed "a three-year moratorium on the holding of any referendum on independence." When the two sides could not agree on conditions, a European Union envoy joined the talks. A framework was agreed in February 2006, and parliament adopted a special referendum law on March 1.

That law set a double threshold. In the observers' words, "55 per cent of the valid votes had to be cast for the 'yes' option, and a voter turnout had to be over 50 per cent of the total number of registered voters." The final report explained the reason: "Such referendum conditions reflected the political consensus within Montenegro that a decision for independence warranted a greater degree of support than a simple majority vote." It added that the 55 percent rule meant the outcome "could not be considered as a foregone conclusion, thus discouraging boycotts."

The question was direct: "Do you want the Republic of Montenegro to be an independent state with full international and legal personality?" The observers found it "clear, ensuring that all voters were able to express their choice without ambiguity."

Both sides had equal places on the referendum commissions and polling boards. An independent chair from Slovakia held a casting vote. Each side received equal public campaign funds.

On May 21, 2006, 484,718 voters were registered, and turnout was 86.5 percent. Final results announced on May 31 showed independence approved by 55.5 percent of valid votes. The observers concluded that "Both criteria for the decision in favour of independence to be considered as valid had thus been met." On June 3, 2006, parliament acknowledged that Montenegro was "an independent state that would seek full international and legal personality." The United Nations admitted Montenegro as a member on June 28, 2006.

What These Cases Show

For the longer record, see Matt Qvortrup's The History of Ethno-National Referendums, 1791-2011 and Aleksandar Pavković and Peter Radan's Creating New States. The official records above point to several lessons about process.

The question decides a great deal. Quebec's 1980 question asked for a mandate to negotiate. Canada later wrote into law that such a question cannot produce a clear result. Scotland's question was tested on voters and shortened. Montenegro's named the outcome in plain words. The strongest questions name independence directly.

The rules should be settled before the campaign. Scotland's vote rested on a written agreement between two governments. Montenegro's threshold came from negotiation between the two sides. In each case the rules were fixed before anyone knew the result.

The threshold is a choice, and it should be made openly. The Scottish and European Union referendum Acts set no turnout floor and no supermajority. Montenegro required 55 percent and a turnout above half the electorate. Canada's Supreme Court called a clear majority "a qualitative evaluation." Whatever the standard, voters should know it before they vote.

Neutral administration builds trust. Hundreds of independent observers watched Scotland's count, according to its Chief Counting Officer. Montenegro gave both sides equal places in its referendum administration and invited international observers. Turnout in all four independence votes was 84.6 percent or higher.

A vote opens a process. Montenegro's parliament acted within two weeks of the vote. The United Kingdom took more than three years to leave. The Supreme Court of Canada's Quebec reference shows that a clear vote creates a duty to negotiate, and Scotland's 2022 case shows that the legal path to a vote can itself be contested.

For Texas, the lesson is that a decision of this weight belongs to the people of Texas, and that how the vote is framed and run will shape how the result is received at home and abroad. A clear question, rules agreed in advance, impartial administration, and open observation are what turned these votes into decisions others accepted.

Sources

Next in The Case of TexasSelf-Determination in International Law

Join the Reading List Donate