Here we explore three applications of referendums. The first is in relation to the framing of new laws, where electorates propose and then vote for specific laws, the outcomes binding on the legislatures. Switzerland and California are notable examples and the Californian system looks particularly attractive. In its so-called “ballot proposition”, a public petition under the “initiative system” can end up with a referendum on a new law.

However, compelling legislatures to frame laws in response to popular demand is problematical. It can create inconsistencies and anomalies within the legal code, and contravene treaties. To an extent, these problems can be avoided by requiring proposals to be compatible with the constitution. A greater handicap, though, is that the process is prone to abuse by well-funded or dedicated single-issue groups, and by the popular press or television. This exposes law-making to rule of the mob. Giving the public direct access to the law-making process can end up in petty tyranny.  As, with official bodies, therefore, we need checks and balances to avoid the system becoming oppressive.

The point here is that there is rarely a problem in getting laws made under the current system.  Largely, the public is able to raise a hue and cry sufficient to force the legislature into action when there is a perceived need for a new law, but the outcome is often bad or ill-conceived law. Contemporary examples include gun laws brought in after Dunblane and dangerous dog legislation initiated after children had been mauled. 

To enable the public voice to be heard, we see no reason why a “take note” referendum should not play a part in raising issues, calling for legislatures to consider new laws. A formal requirement for referendums could even be included in a written constitution. Nevertheless, we believe that such referendums should not be binding.

The second application for referendums we considered was one which addressed a crucial deficiency in our system: the absence of restraint on legislative incontinence. Official systems make too many laws so, rather than making it easier to produce new laws, we need to make it harder. We also need a mechanism to get rid of laws that we do not want.

Support for these ideas comes from Peter Kellner, president of the YouGov polling company. He refers to the great jurist A V Dicey who in 1890 advocated them as a “people’s veto” to block unpopular legislation. Kellner would revive Dicey’s idea, making it possible for the electorate - local or national depending on the issue - to say to their elected politicians: “this time you have gone too far. What you propose is utterly unacceptable”. It would apply to new laws or regulations, after they had been approved but before they had been enacted.

Despite his enthusiasm for them, Kellner felt that political life “would surely seize up if referendums became as common as citizen initiatives in California”. They could act as a deterrent against the abuse of power by, say, a government with a large majority but, on balance, they would be rare events.

One way of ensuring this was to require the signatures of at least one tenth of the relevant electorate before a referendum could be held, within one month of Royal Assent in the case of national laws.  Then, an Act might only be overturned if the majority of all electors, and not just of those who turned out, voted down a measure. That would identify those occasions when Parliament had truly defied the considered wish of the people. Referendums would then occupy their rightful place as a barrier to tyranny when politics fails, but not an obstacle to reasoned reform.

Kellner’s idea of a “people’s veto” – welcome though it is - does not address the “upstream issue” of who holds the power. At the moment, once Bills have gone through all their stages in Parliament, the Monarch gives Royal Assent and they take effect. In this context, the Monarch is nominally sovereign (even if there is no occasion when a reigning Monarch would refuse assent). The people have no say in the matter.

This lack of involvement brings in the second application for referendums. If the people are sovereign, they must be directly involved in law-making. Every law must have the consent of the people and no law should come into force without that consent. Furthermore, this should not be a rare event. It should apply to every law. 

Modified for popular consent, there would be an opportunity for the public to lodge objections to any Act of Parliament, before Royal Assent. Similarly, once an SI had been “laid”, there would be a period during which objections could be lodged. If a threshold level of signatures was reached within a requisite time, there would be a referendum. Otherwise, the law would automatically be deemed approved.

That leaves existing legislation, whether Acts of Parliament or SIs. One idea suggested at Harrogate was that we should adopt the “sunset clause” concept, where laws expire after a defined period unless renewed. But we felt that, if a law is needed (and why else, should it be passed?) it should remain on the statute book. Allocating an arbitrary expiry date, requiring active legislative input to keep a law going, merely adds unnecessary complications and increases workload.

Rather, there should be a specific process. A formal complaint could be raised against and law and, within a prescribed time, should that attract a set number of signatures, a referendum would be held. A majority vote against the law would secure its removal from the statute book.

What applies to national level should, of course, apply locally. Laws made by local sovereign legislatures should have their laws subject to either positive or negative referendum procedures, and there should be provision for removing existing laws. In all cases, referendums would be triggered by smaller numbers of signatures, possibly based on a proportion of the electorate. A figure of ten percent has been discussed.

Then, what applies to laws should also apply to the ratification of treaties. As experience with EU treaties demonstrates, international treaties can be back doors into the statute book, by-passing democratic systems. They become an indirect way of making rules which bind us.

Some treaties, nevertheless, are minor affairs, with little but administrative consequences. Others, such as the Lisbon Treaty, have major constitutional effects. Thus, any proposed EU treaty or treaty change which would transfer powers from the UK to the EU is now subject to a mandatory referendum via the European Union Act 2011.

The most logical way to secure direct democratic approval for such treaties is to adopt the same procedures used for approving new laws. Firstly, there has to be a requirement that no treaty (of any nature) can take effect until it has received popular approval.  Then, those such as EU agreements which alter or add to the constitution would require positive assent, by way of a referendum, as mandated by the European Union Act.

For existing treaties, there should be provision for popular abrogation, although there are complications. Under international law, once a treaty is agreed and ratified, it remains in force unless there is specific provision for expiry. There is no routine provision for ending a treaty by public demand.

Such a system has to change. An ancient privilege, the Crown prerogative, cannot be used to bind and obligate a free people. That it can be so used is evidence that we are not a free people. Agreements with the United Kingdom should only be valid if her peoples are party to them.  Nor is it sufficient that Parliament alone should have the power to decide. Obligations and expense must be borne by the people as a whole. It is, therefore, the people who should have the last word.

Finally, this brings us to the third and last element which would require public approval, permitting challenge and the possibility of rejection.  This category would include certain types of decision made by government or official bodies, - by elected and appointed officials, including ministers and judges.

Clearly, there could not be a referendum for every one (or even a tiny number) of the hundreds of thousands of decisions made each day, so the type of decision amenable to challenge would have to be restricted. Mainly, the “negative” resolution procedure would have to apply, where decisions are deemed to have been approved unless challenged.

With certain types of formal decisions,  such as planning approvals – and even, maybe sentences handed down by judges for certain types of criminal case - one could see referendums triggered by a set number of objections, with a majority vote enabling a decision to be rejected.  In effect, this would be a form of popular judicial review.