Andy Burnham has announced a national commission on electoral reform, and made a commitment to put changing the voting system in Labour’s next general election manifesto.

The implication seems to be that he wants to enact reform without a referendum. That aspiration is understandable. The UK last held a referendum on the voting system in 2011, when the coalition government put the Alternative Vote (AV) system on the ballot paper. As research has shown, reformers can fairly argue that AV was defeated then, at least in part, by widespread misinformation. Referendums don’t have to be like that: notably, New Zealand has done much better. But the scars of 2011 run deep.

At the same time, securing legitimacy for a new electoral system is vital. We do not want the system to become a political football, kicked around by successive governments in response to their particular needs. Something like that has happened in Italy over recent decades, and it harms confidence in democracy.

So far as possible, such fundamental rules of the game as the voting system should be decided not by the players of the game – politicians – and far less by one of the sides in that game – the government of the day – but by the political community as a whole. The standard way for generating consent in a democracy is through elections, based on manifesto commitments. But a manifesto commitment alone would be an insufficient basis for changing the electoral system with legitimacy. This is an issue of the highest importance for our democracy; but it will not be the main factor in most voters’ decision-making when the cost of living, the future of social care, and the security of the country’s borders are also on the ballot paper. Majority victory for a party promising this change would give only a weak signal of public consent to it.

This is why Burnham’s pledge to establish a national commission on the issue is crucial. To confer the legitimacy that any reform would require, such a commission will need to satisfy three conditions.

1. Secure public involvement

First, it will need to be inclusive, bringing members of the public into the discussion as far as possible. The last commission to examine the voting system – the Jenkins Commission established by Tony Blair in 1997 – comprised five members of the great and the good ?. Much as they tried to consult widely, such an exclusive process would not be credible today. It will be important to stimulate national discussion of the issue, and to engage a cross-section of the public in detailed examination of the options.

2. Don’t aim for a particular result

Second, the commission will need to be open-minded as to the outcome it might reach. A process that appears to have been designed to deliver a particular result will not be taken seriously by those who oppose this direction of travel. The status quo will need to be an option considered on equal merits with the alternatives. Those alternatives should include majoritarian options such as AV, a variety of proportional systems (such as list systems and single transferable vote), and intermediate arrangements.

3. Listen to the experts

Third, the commission will want to draw on expert analysis. Advocates of different electoral systems make numerous claims about the effects that different systems would have in the UK – not just upon election results in themselves, but also upon rates of election turnout, the culture of politics, the stability and effectiveness of government, and outcomes such as economic growth and inequality. Such claims will need to be evaluated carefully against all the available evidence. We at the UCL Constitution Unit will be launching a project next week to do just that. Many other experts will have much to say as well. It will be essential to hear a wide range of voices.

A credible proposal is already available

What would a national commission look like that met these design requirements? In fact, a cross-party group of MPs – the All-Party Parliamentary Group for Fair Elections – published a very credible proposal last year. This involved a tripartite structure: an independent commission of seven members who would steer the overall process; an expert panel who would provide “politically impartial advice and evidence”; and a voters’ forum – a citizens’ assembly of around 100 members, who would hear the options, deliberate in depth, and make recommendations.

At least in broad terms, this scheme takes the right approach. Electoral reform is an ideal topic for a citizens’ assembly: indeed, this model of democratic discussion was first devised to examine options for such reform in British Columbia in the early 2000s. The expert panel would allow evidence and analysis to underpin the discussions. The independent commission would have a central role in shepherding the whole process: most importantly, it would need to ensure the process was conducted impartially, so its membership would have be constituted with the widest possible agreement. Besides these elements, the commission would also need to have the means to stimulate a national conversation.

Whether this is the design that Burnham has in mind, we don’t yet know, but something like it will be essential. The electoral system is the very heart of our democracy. Any decisions about it need to be taken with the greatest possible care.