Principles for calling referendums
This thread header is a response to Mr Meeks’ thread of 6 November, in which he wrote, “we can at least set out some principles for calling referendums.” It does not ask whether referendums are desirable (though I believe they are). Instead, I look at how the UK political system should accommodate more direct democracy in the future.
An excellent study by UCL in 2017 noted 550 nationwide referendums in 43 stable democracies between 1990 and 2017, of which 261 took place in Switzerland. Apart from Italy (which held 57), all of the top ten countries were relatively small, with populations of less than ten million. Large countries, such as France (3), the UK (2), the United States (0), Germany (0) and Japan (0), tend to be much less enthusiastic, though this is far from a hard-and-fast rule, as Italy, Belize (0), Jamaica (0) and Israel (0) demonstrate. In addition, large US states, with populations greater than many democratic countries, are some of the most enthusiastic holders of referendums. California held votes on 1,253 statewide propositions between 1910 and 2016 – twice as many as Switzerland has held federal referendums since 1798. Florida has held 398 (1886-2016). This is in addition to large numbers of city and county ballots.
This is therefore an area from which we can learn much from international experience. However, each country’s system is different. Therefore, studying foreign countries may inform the UK’s debate, but it should not determine its answers. There are two particularly relevant features of the UK’s politics which make it exceptional, if not unique: the absence of a codified constitution, and the unusually long tradition of Parliamentary government.
The three big questions we need to answer if we are thinking of introducing more direct democracy in the UK are:
- Should referendums be binding?
- Which issues should be subject to referendums?
- Who should be able to initiate referendums?
Should referendums be binding?
In practice, I think this question is less important than it might seem. The EU referendum clearly showed that even a relatively close victory in an advisory vote hands to the victors an almost unanswerable mandate. However, I favour making future referendums binding, for two reasons. A referendum result which may not be honoured is a democratic absurdity. Also, binding votes should encourage voters to take their votes more seriously than they otherwise might.
Referendums are expensive and time-consuming compared to Parliamentary votes or executive decisions. It therefore seems fairly self-evident that they should be called only on important or controversial issues. If this is granted, we might consider using referendums on:
- Constitutional amendments. Most countries which allow or mandate referendums at all do so on constitutional issues. In many democracies, referendums on constitutional amendments are mandatory, though, oddly, Portugal’s constitution prohibits them. Examples are Austria, (arguably) Canada, Denmark, Estonia, Ireland, Italy, Latvia, Luxembourg, Malta, Switzerland (of course) and every American state except Delaware. This category is not directly relevant to the UK, however, as we do not have a codified constitution. However, in a more normal country, the 2011 vote on the AV system might have fallen into this category.
- Questions of individual rights and conscience, such as abortion, the death penalty, drug legalisation or euthanasia. In Britain, as the UCL study notes, these are generally settled by free votes in Parliament. However, they are often the subject of plebiscites in, e.g. Switzerland, many US states and Ireland.
- Major questions involving national sovereignty. This category might include membership of the EU (eleven out of the thirteen joiners this century have conducted referendums) or NATO (Hungary, Slovakia and Slovenia), or independence and devolution. Even in Britain, recent governments have gone some way towards establishing a convention that these questions (such as Scottish and Welsh devolution or instituting mayors of cities) should be subject to referendums.
- Other policy issues. This is a “none of the above” category. Referendums on these issues are confined to fewer countries than the previous three categories, but are common in Switzerland and many American states.
Who should initiate referendums?
If we conclude that referendums on some or all of the above categories are appropriate, the next question to decide is who should initiate them:
- automatically, as in some countries for constitutional amendments, or in the UK for the (never used) referendum lock on new EU treaties which transfer power;
- by the executive, as in Iceland, which must hold a referendum if the President refuses to sign a law;
- by the legislature, as in the UK’s referendums; or
- by the people through some form of petition, as in New Zealand, the Philippines, Switzerland or many American states.
If we are going to make greater use of referendums, I favour allowing groups of citizens to initiate them. Triggering them automatically risks holding many votes on issues about which few care. Allowing politicians control means that they will only hold votes when they are fairly sure it is to their advantage. This turns referendums into either, as Attlee said, “a device of demagogues and dictators”, or into a device for risk-averse politicians to dodge responsibility for their views. Allowing the people to decide when they are consulted is more democratic, and hence in keeping with the spirit of direct democracy.
As in some countries, once a petition gathers a set number of legitimate signatures, it should trigger an automatic vote. There should be a registration fee to discourage “Boaty McBoatface” type petitions. Then a vote should be held at the next local election date, to keep Brenda from Bristol quiet. The extra expense and effort should be more than offset by a more involved and empowered electorate. Of course we voters will make mistakes, but at least they will be our mistakes.