This is a guest post from Dylan Difford who has recently completed an MA in Politics at the University of Essex, focussing on party and voting systems in Britain and Europe.
A common feature of many Party List PR systems is an electoral threshold – a pre-set bar that parties have to reach if they want to win seats.
The exact level at which the threshold is set varies from country to country. The lowest is in the Netherlands, where a party has to win just 0.67% of the vote nationally to win a seat, and the highest is currently Turkey’s 10% ‘mega-threshold’ (though there are currently plans to reduce this). Most electoral thresholds are, however, somewhere between 3 and 5%.
The main argument in favour of electoral thresholds is that they reduce party system fragmentation – i.e. parties splitting into smaller ones – with countries with higher thresholds generally having fewer significant parties than those with no or low thresholds. This is typically seen as desirable as a higher number of significant parties often leads to government formation becoming more difficult. There is also a worry that, with no threshold, extremist parties can win a few seats and then use their parliamentary presence to gain legitimacy.
These concerns were apparent in the development of Germany’s 5% national threshold in the 1950s. Memories of the instability of the Weimar Republic meant there was some concern when early post-war elections saw ten parties win at least five seats. The installation of the national threshold in 1957 is often regarded as key in making Germany the bastion of moderate multi-partism that it is today.
Thresholds also aim to make sure that national parliaments are composed of parties with a national reach. This can contrast with single-member voting systems, such as First Past the Post, where parties only need to be marginally the most popular in a small area to win a seat. This is particularly apparent in the Indian parliament, where nearly 8% of seats are held by parties who won fewer than 1% of votes.
National, state or constituency?
Where you set the threshold matters. Belgium and Germany both have 5% thresholds – but Belgium’s is set in each constituency, while Germany’s is based on national vote share. This means that Belgian parties like DéFI, who win around 10% of the vote in Brussels but only 2% nationally, can win a few seats in the Belgian parliament. But German parties, like PDS in 2002, were left with no list seats despite winning more than 14% of the vote in five states, as their poor performance elsewhere left them at just 4% nationally.
In multi-tier list systems or mixed-member systems (like the Additional Member System), it is common for the threshold to include ‘get-out’ clauses for parties that achieve some success at the lower level. For instance, New Zealand’s barrier for list seats is 5% of the vote nationwide or one FPTP seat – something the Māori Party took advantage of at the last election, with their victory in Waiariki offsetting their 1% national vote and enabling them to take a list seat as well.
Nationally Imposed Thresholds
Thresholds that are based on national vote shares and which affect seat allocation at a sub-national level can have peculiar effects. The 2013 German federal election is a great example of this – 16% of voters voted for parties that failed to cross the 5% threshold. So while seats were allocated perfectly proportionally between the parties that crossed the threshold, 100% of seats were being allocated based on 84% of votes – distorting things so much that, between June and September 2017, the British House of Commons was more proportional to its most recent election than the German Bundestag!
At the national level, the 5% cut-off is clean and simple, but German list seats are actually allocated at the state level. This can create perverse results such as in the south western state of Baden-Württemberg, where Die Linke were able to win seats on 4.8% of the vote, while the FDP and AfD won no seats on 6.2% and 5.2% respectively. Nationally imposed thresholds mean that the representation of one area will be, in part, determined by votes cast in other areas and can prevent popular, regional voices from being heard.
While thresholds are generally an effective way at reducing party system fragmentation, there are other ways of achieving this. Smaller constituencies create a higher ‘effective’ threshold making it more difficult for smaller parties to win seats. This effect can be seen in countries like Portugal as well as the Scottish and Welsh Parliaments, where moderately multi-party systems have been created without the need for fixed thresholds.
Do thresholds work?
Where the efficacy of thresholds is less clear is when it comes to preventing extremist parties from winning seats. Today, most of Europe’s parliaments are home to at least one party deemed unsavoury by the political mainstream, with right-wing populist parties frequently attaining vote shares to clear even the highest thresholds – statutory or effective.
The surest guard against such parties actually comes from systems like the Single Transferable Vote – whose preferential nature and small constituency size combine to make it very difficult for small, divisive parties to win seats and can penalise larger parties that have limited secondary support.
Though, of course, if a substantive number of people vote for such parties, they should be entitled to some degree of political representation. There is little risk of many of these parties entering government as they are usually subject to a ‘cordon sanitaire’ – whereby major parties simply refuse to work with them.
The same can’t be said of situations in two-party systems under First Past the Post, where extremist factions can take over one of the parties. The logic of First Past the Post limits voters’ and elites’ ability to prevent them from winning power.