This question was posted on behalf of Gerard Lewis, professor, ESPOL University, Guayaquil, Ecuador

The Question

I am writing from Ecuador, where as you may know the votes are being counted after a recent referendum. There is some controversy about the applicability of the absolute majority requirement in this referendum, based on alleged ambiguities in the Ecuadorian constitution.

The referendum contained a series of 10 questions; with options for a separate Yes or No vote for each (it is also relevant to note that voting is compulsory for most of the population). The likely outcome is that for most or all questions, the Yes vote will predominate over the No, but it is unlikely to reach 50% of the total voting population, due to the number of Null or Blank votes (see partial results at http://app2.cne.gob.ec/resultados/Resultadosn.aspx).

My question is: if the Ecuadorian government went ahead and implemented substantial constitutional changes on the basis of a simple (but not absolute) majority, would this be in line with common practice in countries with an established constitutional system for determining referendum votes? I would be grateful for any informed assistance you could provide on this question.

Summary of responses

There is a majority opinion that when amending a constitution through a referendum, an absolute majority is preferable. There are differing views about whether fifty percent plus one of votes cast is sufficient, or if a higher percentage is required for purposes of legitimacy.

A couple of the replies point out that although an absolute majority is the ideal, in this particular  case a simple majority would be better, given that the Ecuadorian constitution does not explicitly require an absolute majority. Retroactively changing the requirements could, it is warned, provoke political unrest.

For purposes of legitimacy, in addition to deciding on a simple or absolute majority, another important factor is whether there is a minimum threshold for voter turn-out in a referendum.

If it is decided that an absolute majority is required, there are still a number of other questions that need to be answered surrounding who will be included when calculating if an absolute majority has been reached, e.g. all votes cast, only valid votes, etc.

For one expert, what really matters is that a constitution should be more difficult to change than a normal law and that the exact way this is achieved through a referendum is less relevant.

Canada is given as an example where referendums have been held requiring an absolute majority and where now a higher ‘approval quorum’ is necessary.

The 2011 UK referendum is an example of a referendum held on the basis of a simple majority.  

Cases where a threshold for turnout was required were Southern Sudan in 2011, Italy and Columbia, while there was no such requirement in the 2011 UK referendum.

Examples of related ACE Articles and Resources

Forum and debate:

External Resources

•    The Canadian Clarity Act

•    The Constitution of Canada: Procedure for Amending Constitution of Canada

•    Engineering Electoral Systems: Possibilities and Pitfalls, An introduction to electoral systems around the world, NIMD (2008)

•    Direct Democracy Handbook, International IDEA (2008)  

Names of Contributors

1.    Manuel Wally

2.    Horacio Boneo

3.    Alexandre Michaud

4.    Mavis Mains

5.    Monte McMurchy

6.    Wim Nusselder

7.    Fredrik Blanck

8.    Khalid Waheed

9.    Atem Oben Henry Ekpeni

10.  Andrew Ellis

Dear Professor Lewis,

It appears that you have to determine the pool of which fifty percent plus one need to be garnered of to constitute an absolute majority.

Without having read the Ecuadorian framework, one would first determine whether the pool is one of all "registered voters legally compelled to vote," or one of all "turned-out votes."  The next step would entail discriminating between "votes cast" and "valid votes cast." The third limb of the test would determinate whether a "blank" vote constitutes a "valid," an "invalid" vote, or a "nullity." You are dealing with a very common problem, as many electoral frameworks fall short of clearly stipulating these distinctions.

If the Constitution lacks a defined quorum in the form of a minimum threshold turn-out (sixty percent in the Southern Sudan Independence referendum), the Ecuadorian Constitutional Court could purposively read together compulsory voting in referenda and the ambiguous absolute majority requirement.

I am not sure whether this helps. Manuel Wally, PhD, LLB (hons)

Dear Gerard,

About ten months ago I completed a document for UNDP Honduras discussing in detail all the requirements and procedures for consitutional changes in all the Latin American countries.  They are two extremely long documents in Spanish.  Since they were the result of a work conducted for UNDP, I hesitate to post them in a public place, but I do not think that there should be any problem in sharing them with a colleague.  You can send me your address to [email protected] and I will be pleased to send them to you.

Dear Mr. Lewis:

In addition to the information provided above, I thought that you may find useful a brief summary of the UK referendum where the question of the threshold was amply discussed (see attached summary which provides the big lines but you could consult on the UK Parliament website for the pros and cons during MP discussions on the issue of threshold re: result no threshold for the referendum).

Note also that Haitians just modified parts of their Constitution but did not proceed by referendum (in case you wanted to research this case?).

regarding a potential future referendum on the separation of Quebec may be of interest (qualified majorities) at the following links:

http://laws-lois.justice.gc.ca/PDF/C-31.8.pdf

Finally, you may wish to consult the Canadian Constitution on mechanisms to change the constitution (while holding a referendum is neither mandatory nor mentioned (i.e. it belongs to legislative assemblies only) there have been conventions requiring a referendum with a yes by 50%+1 of voting electors before any modifications can be enacted (Charlottetown referendum 1992, Quebec referendum 1995):  

http://laws-lois.justice.gc.ca/eng/Const/page-13.html#anchorsc:7-bo-ga:l_V

Hope this can help a little bit,

Alexandre

The referendum that took place in Quebec in 1995, was a motion to decide whether Quebec should secede from Canada.  It was defeated by a very narrow margin of 49.42% "Yes" (to separate), to 50.58% "No", based on plurality of votes cast.   

Given the concern that Quebec could very well have seceded on a very narrow margin, if the percentages were reversed, the parameters for any future such referendums were more precisely legilslated and now the percentage needs to be higher 60+%, to better represent a clear choice. 

In the absence of clear constitutional clauses I would agree that an Ecuadorian court would default to a simple plurality of votes cast.  

In the majority if not all Western Democracies in order to ratify/amend/modify an essential element of a constitution or to impose a fundamental substantive change in terms of parliamentary/congressional practice of governance a simple majority 50%+1 will prove insufficient as common practice seems to suggest at least 2/3rds/66% as being the absolute minimum.

In Canada, not only must 66% of voter respondants support the change but there must be at least a specific mininum level of participation and the provinces must as well be in support according to an already agreed formula which is on record with Elections Canada.

I believe that in order to amend the US consititution both Houses of Congress must be in agreement with at least 66% support as well as at least 2/3rds of The Individual States.

Common prescriptive practice will indicate that 50+1 in order to commit to a fundamental modification in terms of governance will not do.

Monte McMurchy

This reply is posted on behalf of Wim Nusselder, member of the practitioners network

I’m writing from the Netherlands as well as from the experience of the Netherlands Institute for Multiparty Democracy, which runs and funds programmes in 16 countries, including Ecuador.
In the Netherlands we only have (advisory) referendum experience on a local level.
The Dutch constitution requires a 2/3rd majority in sessions of both houses of two consecutive parliaments (i.e. with elections for both in between) for amendments to it.
The essential thing seems to me that a constitution should be (much) more difficult to change than other laws, for the distinction between the two to be useful.
There are many different ways of making that distinction.
Requiring a larger majority and requiring a majority of potential voters rather than actual voters are some of the options indeed.
Exactly what options are chosen seems less relevant than making the difference between constitution and other laws sufficiently large.

You may be interested in our publication from 2008 “Engineering Electoral Systems: Possibilities and Pitfalls, An introduction to electoral systems around the world”.
It can be downloaded from www.nimd.org/document/334/engineering-electoral-systems--possibilities-and-pitfalls-2nd-edition.

With friendly greetings,

Wim Nusselder
coordinator finance at NIMD
[email protected]

To my mind a simple majority should be enough if the constitution in question doesn't explicitly require an absolute or qualified majority. A totally different issue is if it's prudent not to introduce such a requirement for the future.

Refrendum is generaly a consent of General public/Voters to ascertain the opinion on specific issue in an extra ordinary way other than a routine procedure of legislation.

It should ,in my opinion, be based on absolute majority and not on simple majority.

I think it all depends on the participation, if compulsory or not. If only 35 percent of the population takes part in the referendum, then applying simple or absolute majority systems will not still reflect the opinion of all. However, it also depends on the population distribution and voting age.

In any case absolute majority system, 50+, is the best because it gives legitimacy and general acceptance of the decision taken.

Let me elaborate a bit on my position

.

This reply is posted on behalf of Andrew Ellis, Director for Asia - Pacific, International IDEA

There is some info on this question in the Direct Democracy Handbook. What he refers to is a known practice: the terminology to search for within the text is ‘approval quorum’.  Countries which use it in some form include Lithuania, Latvia and Hungary.  Note also the related idea of turnout quorums, requiring the turnout in a referendum to reach a given level for the ‘yes’ to pass, for example in italy and Colombia: however these have two disadvantages – the ability for those opposed to run an abstention rather than ‘no’.  Both turnout and approval quorums can be problematic when the register itself is problematic – as has certainly been the case in Colombia.

I think I have more in my library, but I won’t be able to access that until July.

The referenced handbook on direct democracy can be found by clicking, here

Best

Andrew.