The confusing scramble by Remainers to vote tactically in this General Election has exposed the failings of the First Past the Post (FPTP) system. Tarun Khaitan (Oxford University) explains why the Alternative Vote system could have delivered a clearer signal about Brexit – particularly as it would have discouraged Labour from engaging in strategic ambiguity about it, and forced voters to deliberate on Plan B if their preferred option was not to be realised.
What do the British people want to do with Brexit? Anyone hoping to find answers in the results of tomorrow’s elections is likely to be disappointed. Only a clear majority in vote share (rather than simply seat share) for the Conservatives can possibly be read as a mandate for Brexit; likewise, only a vote share majority for the explicitly anti-Brexit parties (the Liberal Democrats, Greens and others) could indicate a mandate against it. Since no credible pollster considers either of these outcomes to be within the realms of possibility, and Labour voters are hard to read because the party is sitting on the fence, it is unlikely that we will know what the British people really want to do about Brexit on Friday morning. In this post, I will consider which electoral system would have produced the most faithful representation of what the people want. I will be assuming that although Brexit is not the only issue in this campaign, it is clearly the most important one and would largely determine the voting behaviour for most people.
Deciding Brexit through FPTP
The power to deliver, delay, modify, or stop Brexit, on the other hand, will be determined by the relative seat-shares of parties, which—in the existing first-past-the-post system (FPTP)—will likely reflect an over-translated (relative to vote-share) seat-share of the winning party. By doing so, FPTP clearly accrues a significant democratic deficit. Proponents of FPTP claim that this is a price worth paying precisely because it over-translates the votes of the winning party into seats. It is said to deliver stable governments and to exert a defragmenting centripetal force in the political system by penalising smaller parties and rewarding larger ones. It is said to be especially hostile to parties that behave like factions, and try to polarise the electorate by targeting a scapegoat minority.
However, FPTP exerts this centripetal force under very specific circumstances — a unitary state with a relatively homogenous population and a two-party system. Because it relies on single member constituencies, it penalises factions only if social groups are geographically dispersed. In federal states with geographically concentrated ethnic and cultural minorities, FPTP results in a multi-party system with a large number of ethnicity-based or region-specific parties whose political influence is concentrated in their particular regions. India is a good example of a federal system with a heterogenous population and geographically concentrated sub-national groups. In India, FPTP has resulted in numerous factional parties — so while it continues to exercise a centripetal force in each state, and restricts the number of serious contenders to two or three parties, the frontrunners are different in different states. As regional parties like the Scottish National Party, Plaid Cymru, and Sinn Fein gain strength in specific regions of the UK, the only parties that continue to be penalised by FPTP’s centripetalism are the smaller parties with a dispersed support base, such as the Greens, UKIP, and the Liberal Democrats. While this may be welcome to some people, the complete exclusion from power of a party like UKIP is more likely to make it want to seek political possibilities outside the system. It shouldn’t surprise us that parties that are always likely to fail in a system will want to change it, sometimes radically. Whatever its merits might have been 20 years ago, FPTP today simply inflicts costs on UK democracy without delivering its supposed centripetal benefits.
Parties and voters know the democratic deficit in FPTP and have been openly trying to game it. The scale of overt attempts to figure out the best strategy for contesting and voting in these elections has been unprecedented. Given the complexity of the system, and the unreliability of opinion polls, the ability of the voters to successfully game the system is extremely limited. Disgust with being forced to play these strategic games when performing one’s most fundamental civic duty, and frustration with the lack of information that would permit gaming with confidence, has led to renewed calls for a shift to a proportional representation system. Proportional systems, however, are no panacea either.
PR is also problematic
Under a proportional system, a small political party which only seeks a measure of political influence (which can be substantial if it holds the balance of power in a coalition government) may be content with winning (say) 10% of the legislative seats. Under a pure proportional system, it would need 10% of votes to do so. This may be relatively easy to achieve, in many contexts, by running a polarising campaign against a hated ethnocultural minority. In fact, it may even be easier for a small party to secure 10% of the vote share through a distinctive polarising campaign than by competing with broad-church larger parties. Proportional systems, therefore, exert a centrifugal force on the polity. So, while a Brexit-focussed election under PR rules would have permitted the voters to register their preference regarding Brexit directly without needing to game the system, it is likely that smaller parties would have been tempted (even more) to jostle for political distinction by employing divisive and polarising rhetoric. The polity then needs to fix increasingly higher vote share thresholds that a small party must win to qualify for legislative seats — thereby replicating (albeit to a lesser degree) the democratic deficit of the FPTP system that it was supposed to fix in the first place.
At any rate, given Labour’s ambiguous position, it is unlikely that running this election under PR rules would have delivered a clear verdict on Brexit. Even if (say) the clearly pro-Brexit parties won 45% of the vote-share (and seat-share) and the anti-Brexit parties won 30%, Labour’s 25% of seats would have been decisive in determining the way forward. While voters don’t need to (and typically can’t) game the system in a pure PR regime, they can have little idea what the resulting post-election coalition government’s policy is going to be. For all its supposed democratic merits, PR scores low on the accountability of parties to their voters with regard to their manifesto commitments. Remember Nick Clegg? Another democratic problem with PR is that it can vest a very small party with the balance of power — the Theresa May government being held hostage by the DUP is likely to become a very frequent scenario under a PR system.
An AV system could have been the best of both worlds
UK voters rejected a move to the Alternative Vote (AV, also known as Preferential Vote or Ranked-Choice Vote) in a referendum in 2011. That was a mistake. If the current election was being conducted under an AV system, no voter would have been required to game the system.
- An anti-Brexit voter would have ranked her preferred anti-Brexit party as her first choice, perhaps another anti-Brexit party as her second choice, an ambiguous or soft-Brexit party as her third choice, and so on, marking the staunchest hard-Brexit party as her least preferred option.
- A pro-hard-Brexit voter, on the other hand, could have done the same in reverse.
- All voters would have needed to deliberate on their second preference—not necessarily something they liked, but an outcome they could tolerate. This could reduce tribalism in politics.
- A voter who is indifferent to Brexit would have decided her rankings based on their policies in relation to the issues that mattered most to her.
Furthermore, strategic ambiguity would have been a less attractive option for Labour. Benjamin Reilly’s research on the Australian experience shows that AV affects smaller parties differently from FPTP. The latter system denies smaller parties with a dispersed base both power and influence. Under AV, parties like the Greens are able to advise their voters on which larger party to put down as their second preference in return for policy deals reached with such parties. So Labour would have had to enter into some form of pre-election deal with the Greens (and, possibly, even the SNP and the Liberal Democrats) to secure their voters’ second preference. To do so, they may have had to get off the fence. Since the deals are struck before the election, voters know exactly what they are voting for, and are in any case able to ignore their preferred party’s advice regarding their second preference vote.
In addition, AV distinguishes between factional smaller parties like UKIP and non-factional small parties like the Greens: while the Greens could equally figure on any rank for a voter, voters are likely to treat UKIP like Marmite. Some would love it and mark their first preference against it; for others, it is likely to be the over-my-dead-body option. AV therefore rewards broad church parties and penalises factions, unlike PR, which rewards factionalism at the margins of the polity. AV’s centripetalism, however, is more democratic than FPTP’s — it assesses not only the voter’s preference, but also the intensity of her preference and dispreference. Deals between parties are typically made before elections, and are therefore subject to electoral accountability ex ante. Also, while AV also keeps factional parties out of power, unlike FPTP it keeps them within the electoral fold and gives them a stake — albeit a small one — in the system.
Admittedly, AV is harder to explain and administer than some of the alternatives, but tweaks such as limited and optional rankings (which permit — rather than mandate — voters to rank their top two or three candidates) can make the system feasible. The right to rank one’s top three choices is very intuitive to most people. AV has the greatest ability to avoid making gaming demands on voters, and thereby avoid is delegitimising effects on democratic politics.
How this election would have turned out under AV is anyone’s guess. But voters would not have felt compelled to vote for reasons other than their partisan preferences, and could have felt fairly confident that parties would not renege on their key election promises in post-electoral coalitions. For a system under siege, that could only have been a good thing.
This blog post draws upon his draft paper on Political Parties in Constitutional Theory, and from comments by Prof Kim Lane Scheppele and Prof Benjamin Reilly. It represents the views of the author and not those of the Brexit blog, nor LSE.