Everyone agrees that one party government is a bad thing, except those who are trying to create one (like some US Republicans at the moment). But what is the right number of parties? The upper limit for the UK would presumably be 650 (the number of MPs). Where on the spectrum, between 1 and 650 is the sweet spot? It seems inherently unlikely that the correct answer is 2, and the evidence is accumulating about this.
There was a lot of discussion about democracy between, let’s say, the ancient Greeks and the seventeenth century, but little of it was about parties. In fact, ‘party’ was used in early English discussions to describe the danger of factionalism which prevented the pursuit of the public good. Sound familiar? In any case, it was certainly not the natural form of democracy.
Parties developed as a way of managing parliaments not as their basis. The problem they solved was how to create a cohesive government; the answer was by organising policy or ideological voting blocs in the legislature. But these blocs have become the basis on which most of us vote. We choose someone who will join and support our favoured bloc in parliament; not many of us know who we are voting for.
So while parties manage politics, they also sometimes subvert or undermine democratic practices and ideals. Paradoxically, having more of them with an effective voice would help: we need to go up from 2, not down.
Early thinkers about democracy gave much more thought to tyrannts than to parties. They thought democracy gave the power of the people to defend themselves against tyranny. That hope is dashed, of course, by having only one party, which is really the maintenance of oligarchy or monarchy through the ballot box. Two is therefore almost infinitely better than one, despite what standard arithmetic would suggest.
On the other hand, early thinkers also worried that democracy would degenerate into anarchy. A succession of mob reactions to immediate circumstances, the madness of crowds, would drive politics: a system of impulse and instant reaction in which the public good would take a back seat. This is the key problem with a ‘650 party’ solution. Put another way, parties safeguard against political incoherence in democracies that are much larger and more complex than those the early thinkers knew. Athenian democracy was probably managing the wishes of 30-50,000 voters, certainly fewer even than the nearly 80,000 voters in the single constituency of Clacton.
If they are crucial to achieving some kind of coherence they can also corrupt politics. They do this not only literally, by giving access for money, but more subtly by managing the political process and their members for party advantage rather than the public good. They pander to donors, represent vested interests, safeguard their position by distorting public news and information, discipline their members and representatives to vote the party line rather than their conscience and perhaps, most of all, they over-promise to the electorate in order to get the power that allows them to do all this. Their hold over the careers of MPs and on the levers of power makes them a vested interest in themselves, but also organisations through which other vested interests can be expressed.
The Levellers, seventeenth-century English radicals, based their whole argument for democracy on this case about corruption and tyranny: ‘having by wofull experience found the prevalence of corrupt interests powerfully inclining most men once entrusted with authority, to pervert the same to their own domination, and to the prejudice of our Peace and Liberties’. Modern anti-politics sentiment echoes that: machine politicians are all the same and none of them can be trusted with the public good. They promise the moon then line their pockets.
This is the key reason to move further toward 650 than we currently are. Having more than one party guards against tyranny, having more than two spreads the risk of corruption in both the narrow and broader senses.
We already have many more than two parties, of course, but we have an electoral system that drastically reduces the prospects of power of almost all of them. Getting rid of first past the post would break the tendency to duopoly and be a giant step toward limiting the temptations of corruption.
In a reformed system, wannabe-corrupters would have to spread their bets across a wider range of potential players, while knowing that all they could hope for is to influence the thinking of a coalition. This might also ameliorate the disease of over-promising—we the people would vote for a bloc knowing that it would push a particular line, seek to shift debate and policy in the right direction, not on the promise that it would create a new society within a single electoral cycle.
Of course, it is we who should set the rules for this, not the people we pay to play the game on our behalf, but the chance rarely comes without violent upheaval. We are stuck in a system that effectively relies on the players to decide, in-game, on whether VAR works or not. Unsurprisingly the answer of each party depends on what has just happened.
This is why Burnham’s openness to reform of First Past The Post is so important, and his willingness to relax the Whip’s discipline on behalf of the party suggests he really is committed to giving away party power.
We won’t get chances like this often: it’s a moment when we could at a stroke multiply the number of parties with a voice in government. Two is far better than one; but it’s not enough to safeguard the integrity of our system. A number above two is much more likely to preserve us from tyranny and both kinds of corruption, while (as long as it is well below 650) maintaining coherence. The quickest route to that is not to hope for a change in human behaviour, but to limit the power of parties through electoral reform.


Key question is : Do you prefer politicians to compromise BEFORE elections and present electors with a program they plan to implement, or to compromise AFTER elections and implement a program that no-one has explicitly voted for? If the former then first past the post and 2 parties is optimum, if the later then a proportional system and higher number of parties (but even then about 5 is the practical limit of most people’s attention span).