Critics of the “first past the post” electoral rule often complain that it is unfair. It seems unfair that (for example) in the 2019 UK general election the Scottish National Party won 7% of parliamentary seats with only 4% of votes cast across the country, while the Liberal Democrats won 2% of seats with 4% of votes.

So, which electoral system is the fairest of them all?

I submit that there is really no answer to this question, and we would do better to discard it.

Different electoral rules have advantages and disadvantages, but they cannot be neatly summarised into “more fair” or “less fair”. Part of the problem is that the set of political parties and political behaviour we you have is itself (partly) a consequence of the rules of the game. There’s not much point in complaining that the Liberal Democrats are treated unfairly under the UK’s current rules, because if the UK adopted (say) Dutch-style proportional representation, the Liberal Democrats would no longer exist in their current form; the whole party landscape would adapt to look something more like the Netherlands.

For a better way of thinking about constitutional design I commend to you three points for a win. Historically, football leagues had awarded zero points for a loss, one point for a draw, and two points for a win. In 1981 the English Football League changed their rules to give three points for a win. After this was adopted for the group stages of the 1994 World Cup, it spread across the world. So, which is fairer, two points for a win or three points? My hope is that transferring this question to football helps bring out its essential absurdity. There just is no “really” fair number of points for a win. And that’s not how the League approached the question. Instead, they started from the point that the rules influence the way the players behave. This makes it hard to think about the rules in terms of fair rewards, but it means rule-makers have a lever with which to change the game. The reason the English Football League started awarding an extra point for a win was because they hoped that doing so would incentivise teams to play more attacking football rather than playing for boring draws. In other words, they started out from the point of the whole thing, which was to produce a good spectacle, and they worked backwards from there to think about how they could change the rules to promote it.

Constitutional design should proceed similarly. Start from the goals of politics, like peace, justice and prosperity. Then think about whether reforms are likely to promote those goals or not. This gives rise to a whole new set of questions when considering political reforms such as changing the electoral system: Will this lead to more parties or fewer? Will parties be more stable or more fragile? Will that make it easier or harder for voters to cast informed votes? More dramatic swings in policy or more stability? More polarisation? More apathy? More geographically-based parties or more ideologically-based parties? Which of these things do we actually want, and which of these things should we try to avoid? These are difficult questions. But they are meaningful questions that are worth asking.

Has three points for a win actually led to more attacking football? I leave that for more competent judges. Whatever the answer, it is the right way to approach the question.

I am a lecturer at Nottingham Trent University. I have worked on the justification of democracy, the relationship between democracy and the market, and the political theory of business corporations.