Are majoritarian or proportional electoral systems better?

This essay was written as a formative part of my undergraduate studies at the London School of Economics.

Electoral systems – the set of rules that regulate competition between parties and/or candidates during elections, that decide how vote shares map to seats in parliament and indeed, how the electorate express their preferences – have traditionally fallen into two categories; majoritarian – which include Single Member Plurality (or ‘First-Past-The-Post’), the Two-Round System and Alternative Vote – or proportional – like open or closed-list PR and the Single Transferable Vote¹. 

This essay will assess the consequences of each type of system, with relation to the formation of parliaments arising from each, leading to a conclusion that neither is more beneficial and in fact a third way is possible and preferable.

The debate in this field is diverse, with many scholars sitting on either side of the proverbial fence on the matter. The debate is heated and often exaggerative in nature; Lewis comments that “the surest way to kill the idea of democracy in a plural society is to adopt […] first-past-the-post…” (1965, p. 71)², whilst Rokkan claims that it is “simply impossible” to make statements about the benefits of proportional electoral systems compared with majoritarian electoral systems (1970, p. 166)³. None-the-less, countries continue to adopt different types of electoral system in the hope of achieving the best political outcomes – but what are these outcomes, and are they a benefit to the citizens within those countries? Traditional argument tends to follow that choosing an electoral system is about an ‘opportunity cost’ or ‘trade-off’; you have to choose between parliamentary representation or government accountability, between cohesive parties or individually accountable politicians, between majority or proportional electoral systems⁴. The suggestion here is that there is a linear trade-off between these two factors and they are mutually exclusive. Of course the crux of the issue with electoral systems is not the system itself, but the types of parliament and government that form because of the electoral system. Proportional electoral systems tend to produce minority or coalition governments, which will produce a very different set of political outcomes than majoritarian systems, which tend to create single-party governments⁵.

There are many perceived benefits to a proportional electoral system; representation gains are among the biggest of these. Because seats are allocated in direct proportion to the share of the vote, parliaments tend to be a microcosm of society, representing a broad range of interests, ethnic backgrounds, economic backgrounds, and so on. The principle outcome is that parliament maps, as closely as possible, the preferences of the median voter and the government that forms, usually a coalition of two or more parties, will do the same⁶; the policy outcomes therefore, are most beneficial to the greatest number of people. This breadth is not witnessed in majoritarian plurality systems; Duverger’s Law showed that majoritarian systems were far more likely to create two-party systems than proportional systems which generated multi-party races (Duverger, 1959)⁷. Having a two-party system does not necessarily mean that the preferences of the median voter are unaccounted for however, as in majoritarian systems, parties will shift towards the median voter in order to gain a larger vote share, and can be sufficiently broad to capture all of the political spectrum. The majoritarian electoral system and two-horse race can mean that representation is disproportionate in parliament, and in many cases the government can have a large overall majority even though they receive less than 50% of the vote. We see this in the UK, where with only 43.2% of the vote in 1997, the Labour party secured a landslide majority with over 63% of the seats in Parliament. Clearly under majoritarian systems, the preferences of voters are not expressed well enough.

On top of these representation benefits associated with proportional systems, there is a by-product in countries that suffer from long-running ethnic or religious conflict, like Northern Ireland, where a proportional electoral system – in this case STV – forces a power-sharing agreement in Stormont; parties have to overcome traditional religious divides and work together. With the exception of some dissident terrorist attacks, the peace accord in Northern Ireland is strong, arguably because of the choice of proportional electoral system that means citizens on both sides of the religious divide have representation within the Assembly.

There are consequences of proportional systems that might be less desirable however. Coalitions, by definition, require compromise, negotiation and debate to survive. If parliaments are formed of many small parties, rather than two large parties like in a majoritarian system, you can end up with a stalemate in government formation. Countries can go for months without a government – like in Belgium, where 200 days after their most recent election, a coalition agreement still could not be reached – but more than this, when coalitions form, they can be made of such divergent ideologies that they are incapable of agreeing on anything, with one partner vetoing the coalitions actions if the policy option is not within their ‘winset’ (Tsebelis, 2002)⁸. Policy inaction, particularly when exogenous shocks occur, is a very dangerous thing; perhaps making majoritarian systems a better option.

The principle gain from majoritarian systems is government accountability because they tend to lead to single-party government. In a proportional system, coalitions formed of several parties have very little clarity of responsibility; governments are generally formed after elections and the policy outcomes are generally very different from those in the coalition partners’ manifestos.  When these policies are implemented and effects are felt, it is very difficult to know which parties to blame and which parties to reward and so, come the next election, poorly performing parties from the coalition will likely remain in power. In effect, proportional electoral systems can serve to entrench bad government permanently. Of course, the extent to which this occurs will depend on how parties approach the coalition process. If parties are following an office-seeking theory⁹, where policy will be very flexible so long as parties get in to power then the clarity of responsibility will be low; you could argue that this is exactly the situation in the UK with the current coalition government – albeit that this was formed by a majoritarian electoral system. When 80% of the electorate were against an increase in higher education tuition fees, why did the coalition, which should more accurately represent the median voter, vote to increase fees, and who, of the Liberal Democrats or the Conservatives, are voters to blame at the ballot box? At least, some would argue, under a majoritarian electoral system, you are more likely to get single party government and you are therefore, as a voter, more easily able to identify who is responsible for the policies coming from parliament. Furthermore, if the government is doing a bad job, a small shift in votes will equate to a radical shift in seat allocations in Parliament – though in countries like the UK, this means elections are fought in a minority of ‘swing seats’ and the rest of the country is not engaged fully in the electoral process, which is clearly a negative consequence. Another consequence of single party government and large seat swings is that you are even less likely to have policy that maps to the median voter; you are likely to have a skew, either to the left or right, and so again the trade off arises – do we prefer policy that maps our interest more closely – and so do we prefer proportional systems – or governments that we can attribute reward and punishment to – and so prefer majoritarian systems? Many political scientists would argue the former is preferable, notably Lijphart who argued that proportional systems were “virtually synonymous with electoral justice” (1984, p. 140)¹⁰.

But there is a third way. Several political scientists have suggested that rather than a direct trade off between these factors that the relationship is actually able to be maximised in a way not dissimilar from that shown in  figure 1. This can be achieved by creating small multi-member constituencies (Shugart & Wattenberg, 2003)¹¹. The system works because, principally, it decreases the incentive to ‘tacitically’ vote (Cox, 1997)¹², it reduces the number of people able to co-ordinate around individual candidates and allow for a broad range of elected representatives whilst still maintaining a line of accountability. Both accountability and representation are desirable in parliaments and neither a pure majoritarian system, nor a pure proportional system can provide both. A small multi-member mixed electoral system however can reduce the unrepresentativeness of parliaments by 75% and reduces the ideological distance between the median voter and the government more so, whilst it only increases the average number of parties in government by one-half and adds one large viable party to parliament, and though coalition is more likely, it only includes two or three parties at most, not the same sort of fractious and broad coalitions that we see in Belgium and Iraq. (Carey & Hix, 2010)¹³. Clearly a system that can give us the “best of both worlds” is the best option.

In conclusion, it is clear that in reality neither electoral system – majoritarian or proportional – will be better than the other, because neither give us the desired political outcomes we desire; they perpetuate the idea of a trade off. We should in fact seek a middle ground; a well designed electoral system should be able to combine the best of both worlds – high accountability and fair representation (Carey & Hix, 2010)¹⁴. A mixed multi-member system with low district magnitude provides this, and I would suggest to you that the middle option is the best.