Delegated Vote

The Delegated Vote has a very different approach to the role of constituencies. The top two candidates are automatically given seats with different voting powers. ¹ The remainder of the vote is dealt with by merging constituencies. This system allows us to retain constituencies, (essential for maintaining local representation) whilst allowing parties with a fraction of the national vote to gain representation.

Although this system uses constituencies, it is not bound by them. The universal value of the vote means that votes can move safely between constituencies. Candidates stand in more than one constituency and pool the vote between them. To gain a seat they need to have received enough votes to put them above the average of the bottom fifty constituency MPs but also below the average of the top fifty constituency MPs. The lower limit exists to ensure that although there is variation in the power of MPs, there is no difference between MPs elected by single constituencies or those elected by multiple ones. The upper limit is there to ensure that it is never to anyone’s advantage to concentrate more votes in fewer individuals. This can loosely be seen as Gerrymandering rehabilitated, instead of drawing boundaries to bury opposition parties as minorities parties can effectively draw their own constituency boundaries so their likely voters are inside without being able to exclude or bury the voters of other parties.

This is the aspect of the system that needs most development as it requires parties to take an active role in creating proportionality. This is a disadvantage it shares with other multi member systems, STV requires parties to create proportionality by standing the right amount of candidates in the constituency (too low and they risk not gaining all the votes they count, too high they risk loses votes in second preferences). However, whilst similar - the challenges faced here are undeniably harder as they require information gathering and action in a wider area.

The party must create a constituency for itself by merging existing constituencies until it has enough votes. As this must be done before the election so the candidate can be corrected named on the ballot, this relies on polling. However, this has added complexity as because the upper and lower bounds are set by the votes received by constituency MPs, they must try to hit between moving targets².

For small parties this should be relatively straightforward, accumulating small amounts of votes in each constituency should allow for fairly predictable sizes. The main problem is for parties where it is unsure if they are a second or third place contender. This is the only point in the process that one party is concerned by the electoral placing of another as if more than two parties only stand in a single constituency, one party will lose out. To resolve this requires inter-party agreement where parties would agree which two parties stood as the single-constituency candidates. As this is likely to be problematic, it would make sense to deed the responsibility to an independent body. A new role of the Electoral Commission would be to conduct the polling to determine the location of potential votes and advise parties on placement of candidates. It could balance across the country which parties it placed into the single-constituency positions far more fairly than parties are likely to achieve on their own.

Using MPs with different voting powers lets us use constituencies of different sizes. Whilst most constituencies must be of roughly similar sizes for the upper and lower boundaries to have any meaning, we can create exceptions for constituencies that already have far less people than the average and whose sheer size makes further enlargement impractical³. Another possible use would be to create different groups of constituencies with different average sizes and upper and lower boundaries for multi-constituency candidates. This would allow for constituency sizes to be different in urban and rural areas for example. This would have the disadvantage of creating an urban/rural MP divide with more rural MPs (but with less power) and would prevent votes moving between the two groups. The lower population sizes DV requires in comparison to other multi-member systems should reduce the need to divide into different classes of constituency, but it remains a possible use of the system.

1. This is a number open to variation; the goal is a large proportion of votes will be represented locally. At least two are required to give local choice, it may work better with more (but this would increase the size of constituencies). [↩]
2. I suspect these bounds will rarely change by enough that this is an issue and there will in reality be a rough range to aim for, but this is an area that needs to be investigated. Potentially the boundaries from the previous election could be used if it proved they changed too much between elections to be predicted. [↩]
3. These would not be counted when creating the upper and lower boundaries for multi-constituency candidates [↩]