It might well be that the United Kingdom, or its successor rump state of England and Wales, will be relying on the skills of the New Zealand’s trade negotiators to help shape the Brexit agreement with the EU. Amusingly, these might be the same people who are also representing  New Zealand and Australia in areas where those two countries collaborate to reach common terms with the post-EU British / English-Welsh state.

That’s a mouthful of a paragraph because it’s a mind-blowing idea, or should be.

But it would unlikely to have become reality had it been thought about before the Brexit referendum.

Unfortunately, we have somehow got it into our heads that referenda are binary, yes / no questions.

But they needn’t be.

And we could have learned that lesson from New Zealand before forcing many people to choose between the status quo and an option that was, really, many options, none even remotely defined.

Last year and this, New Zealanders voted in two referenda designed to address one issue: to keep the current flag or replace it with a different design.

In developing the question to be put to the electorate, prime Minister John Key, his advisors and the parliamentary committee tasked with establishing the rules under which the referendum would happen realised that a simple yes / no option along the lines of “would you like to replace the current flag of New Zealand with a new design?” might well have resulted in a yes vote. There would then have followed a lengthy period of bitter argument about what the resulting flag should look like, at the end of which a significant percentage of the population who had voted for change might well have ended up wishing after seeing the new flag that they had voted, instead, to keep the current one.

Thus, in late 2015 a first referendum asked voters to choose between various possible designs for a new flag. The winning design then went up against the existing flag in a second referendum, which asked voters to choose not between an existing reality and hypothetical future but between the status quo and a specific alternative.

And so the government is left, at best, trying to negotiate an exit that is guaranteed to be opposed by more than 50 percent of those who voted in the referendum [the assumption here, of course, is that Remain voters would have chosen that option regardless of the proposed alternative—likely a safe bet].

There is lots of talk / vitriolic shouting about “democracy” at the moment. Much of it is valid. But democracy that works seldom works as a series of binary yes / no choices. Sure, there are occasions when matters are that simple, but usually some element of nuance is required to make an informed decision.

As the Liberal Democrats continue to develop what is rapidly becoming the most coherent, principled and thoughtful response to the Brexit vote, perhaps the party should also take on board a commitment to a referendum process that uses a two-step mechanism to ensure voters are clear about what they are voting on and in which they have been involved in the process of refining the alternatives.

And, yes, it does need to be a two-step referendum. And, yes, those questions need to be asked separately.

On November 6, 2012 Puerto Rico’s residents got to try their hand at making sense of a two-part question that just about guaranteed no one could understand the outcome. Residents were asked on the same ballot whether Puerto Rico should retain its current status with respect to the United States, a majority (54 percent) voted No. Part two asked voters to choose between statehood, free association with the United States, and independence, a set of choices that ignored the status quo option. The statehood option got 61.6 percent of the vote , but, and here is the crucial point, a large majority of the people who voted for no change in status simply left the second question blank. After all, their choice was not an option in the second part. Given the split in the choices among those who had voted for change, the blank ballots on part two were the second largest number of ballots and, if counted, would have meant that no option got 50 percent of the popular vote.

In this case, the design of the referendum meant that while plenty of choices were available it was all but guaranteed that no one option would emerge as a winning position and that almost no one would find the result persuasive.

It’s not that we should try to complicate politics unnecessarily, but sometimes two choices are not enough, and three or more in one go can’t possibly work.

Think of a two-step referendum as something akin to preferential voting and it makes a lot of sense.

* Chris Fauske, an Essex-native, is now is a resident of Massachusetts.