What kind of democracy do we want?

Until now, Britain’s voters have normally chosen not just their MPs but, in practice, their prime minister. Those days may be coming to an end

Two weeks ago I argued that we need to change our voting system: in general because the gulf between shares of votes and shares of seats in last year’s general election was indefensibly wide – and specifically because there is a danger that in the near future Britain could have a government that most voters actively oppose. At the end of my analysis I promised (threatened if you prefer) to discuss the options in a future post. Here goes.

Long-standing advocates of proportional voting generally make a simple, compelling argument. Each party’s share of the seats in the House of Commons should match its share of votes across the United Kingdom.

My problem with this is not that it is a bad objective, but that it’s not the only one. General elections have three functions: to link votes in the country to seats in Westminster; but also to decide who should govern Britain; and to give voters local MPs who represent their interests and concerns. Last year’s election failed the first of those three tests, but passed the other two.

Sadly, no system scores 100 per cent on all three tests. What we should do depends on the relative importance we attach to each objective. This is something on which reasonable people can disagree. (Indeed, some people may want to add other objectives, such as the ethnic and gender balance of our MPs.)

For example, the Single Transferable Vote (STV) in multi-member constituencies (used for national elections in Ireland and Malta; and in local government elections in Scotland and Northern Ireland) is good at producing proportional outcomes for all bar the smallest parties. If we were to divide the UK into, say, 130 five-member constituencies, each voter would be able to call on any one of five “local” MPs – but these would typically represent half a million citizens, rather than be a single MP representing a population of around 100,000. Is that satisfactory? There is no right or wrong answer. It depends on how local we want our MPs to be.

The other main way to generate proportional outcomes is the Additional Member System (AMS), used, for example, in elections to Germany’s Bundestag and Scotland’s parliament. Citizens have two votes, one for a local representative, using first-past-the-post (FPTP), the rest for a party list for their region. The second, list, vote determines how many seats each party should have.

In the last Scottish elections, the SNP’s list votes entitled it to 64 seats in the 129-seat Holyrood parliament. As it won 62 local seats, it was awarded just two list seats. The Conservatives’ votes entitled it to 31 seats overall; as it won only five local constituencies, it picked up 26 regional list seats.

If we were to adopt AMS for Westminster with a 50-50 split between local and list MPs, then each constituency would be double today’s size – typically 200,000 rather than 100,000.

There is one more key feature about different voting systems. Many voters adapt their choice of party to the system they have to operate. Last year’s tactical voting was a consequence of FPTP. Many anti-Conservatives voted Labour or Liberal Democrat depending on which was better placed locally to defeat the Tories.

What’s more, smaller parties such as the Greens often suffer under FPTP because people don’t want to waste their vote locally on a candidate who can’t win. A proportional voting system would help them. Indeed, it could well give parliament more medium-sized party groups (say, 50 to 200) at Westminster and make larger parties a rarity.

So much for the arithmetic. How about the politics? The most obvious consequence of adopting either STV or AMS in Britain today is that there is little chance of any subsequent election producing a single-party majority government. Sometimes even two parties wouldn’t be enough. Multi-party governments would become the norm (instead of the exception as in 2010) and often need three or more parties to work together.

Would that be a good thing? Again, there is no right or wrong answer.

Advocates say that a) a proportional system is right in principle; b) for the first time in recent history, we would have a government made up of parties who together won the votes of a majority of all voters; c) multi-party coalitions abroad have tended to lead to more successful administrations than Britain; and d) there would still be adequate local representation, even if each MP had to serve many more people.

Each of those points can be debated at length. For example, do countries across Europe do better when they are constantly governed by multi-party coalitions? How do we judge “better”? To what extent are national performances caused by their voting system rather than their history, culture and geography? Define your terms and pick your evidence, but please don’t claim that the answers are beyond dispute.

However, the basic issue is one of democratic principle. It concerns the relationship between the executive and the electorate. When party competition is fragmented, proportional systems transfer the ability to choose the government from voters to MPs. (Yes, technically, British governments depend on having as majority in parliament; but in practice we normally know who will be prime minister as soon as the votes are counted.)

Fundamentally, the case for FPTP is that its bias towards larger parties has generally given voters a clear choice of government at election time. The cost has been that no incoming prime minister on modern times has been able to claim the positive support of a majority of voters. Proportional systems generally ensure majority rule (in terms of the total votes won by the coalition parties) – but with governments resulting from often tortuous post-election negotiations.

The one clearcut example of coalition government in Britain illustrates the problem of defining “majority rule”. The two parties in the Cameron-Clegg coalition secured the votes of 59 per cent of all UK voters (Con 36 per cent, Lib Dem 23 per cent). But how many Lib Dem voters wanted a Tory-led government? How many Tories wanted to join forces with the Lib Dems?

For most of my life the killer argument in favour of FPTP has been that elections usually gave voters the most important democratic right: to kick the rascals out and instal the opposition. As I argued two weeks ago, the clarity of choice this has successfully trumped the fact that winning parties never reached 50 per cent support. When power has changed hands at Westminster, it has broadly been what voters wanted – even last year. To be sure, only one in three voters backed Labour; but the vast majority wanted the Tories out and were relieved when Keir Starmer took over the next day.

Proportional systems can complicate the process of removing an unpopular government and installing an alternative. Again, consider our 2010 election. If the same votes had been cast under a proportional system, a majority government could have been formed by any two of the three main parties. If 59 per cent could be said to support the Con-Lib Dem coalition, by the same token, 52 per cent would have backed a Lab-Lib Dem coalition (Lab 29 per cent, Lib Dem 23). Sixty-five per cent voted Tory or Labour. Would that have been the most popular coalition of all? Similar three-way options for two-party coalitions could have applied to 13 of the 21 general elections held since 1950.

However, while FPTP has generally saved us from these quandaries in the past, it may not do so in the future. It might no longer offer voters a clear choice of government – or do so, as last year, at the price of converting votes into seats in a way that a democrat finds hard to defend. If we are drifting into an era where governments are routinely negotiated after the election rather than decided on election night, then the main reason to keep the present system, the ability to kick the rascals out, falls away.

One way to look at the way forward is this. As I argued last week, Reform could secure a majority at the next election with as little as 29 per cent of the popular vote. Right now, I would expect the party to slip back before the next election, and not be a contender for power. But I can’t be certain, and nor can anyone. Which would we prefer: FPTP and a small chance of Nigel Farage leading a majority government – or a proportional system and a bigger chance of him joining a post-election coalition?

Just now, this is not a live political issue. It should be. Instead of leaving it to an argument behind closed doors after a messy election result by people looking only for partisan advantage, wouldn’t it be nice to have a calm national debate about the kind of democracy we want?

Well, we can dream.

So, no system is perfect. My third and final post in this series will suggest the best available way to balance the various objectives of general elections.