CONVENTIONAL wisdom since the 18th century has insisted that modern democracy be representative in nature, structured through competitive partisan elections. Outside the pages of utopia, it was assumed that the people – although the ultimate source of authority – were too ill-informed, too prejudiced, too flighty, too irresponsible, to be given more than an indirect and occasional voice in governing.

Citizen’s assemblies challenge that assumption. They bring together randomly selected ordinary citizens to decide matters of public importance through informed deliberation.

Ordinary people decide, but not ignorantly or unreflectively. They decide only after having heard expert witnesses, considered the question in all its aspects, and deliberated amongst themselves.

In Canada, the Netherlands, Iceland and Ireland, citizens’ assemblies (and similar bodies) have shown that people from all walks of life, socio-economic classes, and educational levels are willing and able to get stuck in – to question experts, to disagree with civility, and often to reach sensible compromises.

Participants in a citizens’ assembly generally find it a positive experience. Former participants are likely to become, and to remain, more politically informed and active.

Citizens’ assemblies are both a great vindication of democracy’s faith in people’s goodwill and good sense, and a great spur to the development of a healthy democratic ethos.

However, the record of using citizens’ assemblies to achieve constitutional change is mixed. Beautifully inclusive processes have sometimes had frustratingly inconclusive outcomes.

In Canada, the citizens’ assemblies of Ontario and British Colombia each developed good schemes for electoral reform, only for these to be rejected by the people in a referendum. The general public, who had not taken part in the assembly’s discussions, fell back on tired old arguments and voted to keep First Past the Post.

The most successful citizen’s assembly is that of Ireland. The Irish citizens’ assembly is solely composed of ordinary citizens, randomly selected in way that is weighted by class, gender, age and region.

Of the five topics on its agenda three (abortion, the conduct of referendums and fixed term parliaments) are constitutional in nature. On the first and most controversial of these topics, the citizens’ assembly’s proposals were acted upon by the government, passed by parliament, and adopted by referendum.

It is to Ireland that some forward-looking minds in Scotland now turn in search of a solution to our constitutional future. Could an independent Scotland’s new constitution be written by a citizens’ assembly?

The answer is both an emphatic “yes” and a cautious “no”.

Yes, if we mean “Could a citizens’ assembly – or series of citizens’ assemblies – have a useful role to play in developing a new written constitution for an independent Scotland?’

No, not if we expect a citizens’ assembly to do all the heavy lifting alone.

Constitution-building is a process. It needs careful planning and process management. A citizens’ assembly cannot be an amateurish affair. It cannot be the voice of one party. To do its job, it needs broad political commitment at the highest levels.

In Ireland, the citizens’ assembly was established by resolutions of both houses. The chair is a retired judge, appointed by the cabinet. It has a secretariat seconded from the civil service. These details are important. They embed the citizens’ assembly in a supportive institutional environment that make its recommendations harder to ignore.

Workable constitutions must be politically negotiated and agreed amongst political elites as well as amongst the general public. We might not like that, it may cramp our style, but there it is.

The penalty for ignoring existing elites is to have them use their power to crush the process. This was amply demonstrated in Iceland, where an attempt to remove the politicians from the hyper-democratic ‘crowd-sourced’ constitutional process led to distrust and motivated the politicians to scupper it.

Finally, constitution-building requires complex compromises. Throwing things out to a citizens’ assembly and expecting them to find answers may suffice when there is an established and well-working constitution (as in Ireland), which needs only one or two specific tweaks. It is unlikely to suffice when the whole constitutional order is up for grabs.

It comes down to this: there must be room not only for citizens’ deliberation and expert advice, but also for political leadership, consensus-seeking and negotiation. If we are looking to Ireland, a better model may be that of the previous constitutional convention (2012-2014), in which two-thirds of the members were ordinary citizens, but one-third were politicians representing the various parties.

Citizens’ assemblies are a valuable new addition the constitution-builder’s toolkit. They are well suited to resolving specific, deadlocked, stand-alone issues. They can be agenda-setters. But they are not an answer in themselves. A citizens’ assembly must be integrated into the process as a whole, in a way that supports – not replaces – other forums and institutions of constitutional decision-making.

This unique column welcomes questions from readers