Read part 1 here.
The proportionality of representation
Voting behaviour is a complex matter, but over the past few decades there is evidence of a general drift away from voting largely on the basis of party loyalty. Instead, many voters prefer to examine where candidates stand on issues of importance to them and to vote for those closest to their own position.
If that is what is happening with those voters not yet entirely disillusioned with our political system, then we need to revisit the notion that ‘proportionality’ in electoral systems means that the seats gained by a political party are proportional to the votes it received. With the shift of focus from party to candidate, ‘proportionality’ relates seats gained to coherent bundles of policy in a multidimensional political space.
Pessimists will argue that this makes debate and decision-making harder, and they are probably right. But optimists will claim that this shift greatly enhances the scope for representing voters’ wishes, since a closer match is more likely to occur between a voter’s views and those of a candidate than the blandishments of a political party manifesto.
This provides a first clue as to what system of PR might best support such a situation, since we have already noted in Part 1 that, compared with list systems, STV favours proportionality of policies rather than parties.
But there is a second assumption hidden in the conventional understanding of ‘proportionality’, which is that all representatives have an equal say in an assembly vote. The fact that one candidate was elected with, say, 40,000 votes and another with 20,000, is regarded as irrelevant. But is not the number of votes received by a candidate also a factor in proportionality? Does taking them into account not improve the degree of representation of voters’ wishes? Recent proposals for a modified form of STV claim that this would indeed be the case.
Weighting for a change
Guy Major and Jonathan Preminger are Senior Lecturers at Cardiff University. They are the authors of a recently-published paper entitled Democratising Democracy: Votes-Weighted Representation. (Note that quotes in this article arise either from this paper, or from an interview with the authors conducted online.)
After laying out the premise that democracy is in global retreat, the authors summarise a modification to the STV system of proportional representation which, while simple in itself, promises profound improvements in an electoral system’s democratic credentials.
Coining the term “Votes-Weighted Representation” (VWR), the authors suggest that, instead of transferring an elected representative’s ‘spare’ votes – those received over and above the quota needed – to next-preferences as in standard STV, those votes are retained by the representative. Thus (simplifying the earlier example by dividing their votes by 1000), we’ll give representative A a simple ‘weight’ of 40 and representative B a weight of 20. These weights then count in an assembly vote, giving a more precise degree of proportional representation of voters’ wishes.
In their paper, Major and Preminger note that weighted voting is nothing new. Shareholder’s votes at company shareholder meetings, for instance, often carry the ‘weight’ of the number of shares the shareholder owns. And weighted voting is used in the European Council, and has been proposed for use in the UN General Assembly, to avoid giving smaller countries too much voting power. But, stresses Preminger, under VWR, there is a crucial distinction to be made between a shareholder’s vote weight and a citizen’s:
“The shareholder with all those shares and a lot of power because of those shares, is voting on his opinion. Whereas the representative with all those votes is, at least in theory, voting on the opinions of the people who gave him that weight.”
The authors also stress that, while, as a system, this may all sound somewhat complex, it is important to take the voter’s point of view into account too:
“From the voter’s point of view it’s actually a very simple process, even though from the internal mechanism of it, it may seem less simple.”
VWR is, in fact, not that complex at all. It is simpler than standard STV because, while unelected candidates’ votes are still transferred to next preferences, elected candidates’ spare votes don’t need to be transferred, considerably simplifying the vote counting process. And while not suggesting that such vote-weighting is sufficient in itself, the authors claim that alongside measures like including democracy in secondary education, lowering the voting age, decentralisation and reforms of the economy and the workplace, it could be a game-changer.
Is the party over?
In addition to escaping from the iniquities of FPTP, such a reform of the voting system would help to tackle a number of problems arising from our traditional party-based approach, say the authors. For instance, ‘Party-capture’ occurs when elements within a party gain disproportionate control over policy and candidate selection. One doesn’t need a spyglass to see this occurring in both our major parties, even as we speak.
In an extreme form they refer to as ‘leaderism’, the leader of a governing party and their immediate advisers can abuse the power vested in them. For example, they can exert undue control over policy (proroguing of parliament), extend patronage to cronies (honours lists), spend public money indiscriminately without due diligence (Covid PPE), and interfere unduly in candidate selection, to mention but a few.
As with standard STV, the system requires multi-member constituencies (Major and Preminger suggest 6 or 7) of sufficient size to ensure greater heterogeneity of political views, and the authors work through several examples of how VWR would work in these. In a plausible example, they show how 95% of voters get a first-preference representative elected, while the 5% transferred are likely to be to representatives with a close match to their policy preferences.
Within a multi-member district of 6 or 7, the authors claim that VWR should work well, irrespective of the number of districts and any disparities in the relative size of their populations, their geographical spread and their exact boundaries:
“Weighted voting in the assembly can allow different constituencies to have significantly different populations, if wanted: representatives from a lower population district would have fewer votes, on average: weighting-by-votes would ensure approximate overall proportionality of party voting power in the assembly, irrespective of the different populations of different constituencies.”
As well as reducing political asymmetry, VWR, they say, would mitigate against gerrymandering:
“The proposed system is also inherently resistant to gerrymandering, as the weighting-by-votes of multiple representatives per constituency makes overall party voting power in the assembly relatively insensitive to boundary changes. A party might lose some votes in one district but regain more or less the same number in the neighbouring district, if the boundary was shifted into the first district: most of the ‘moved’ voters would in effect simply shift representatives within a party, while preserving its overall voting power in the assembly.”
The advantages of VWR also extend to political parties. They can make themselves more locally appealing by listing different candidates first in their party group in different voting districts to reflect any local ties or to establish one. Major says this makes it easy to select candidates from within a party group if that’s what the voter wants to do:
“You can home-in pretty quickly on the party you’re interested in. You can have their candidates blocked together, for example, rather than listed in alphabetical or random order.”
But more importantly, by providing a range of diverse candidates to choose from, VWR can enable better voter-representative dynamics (the so-called ‘local link’). A voter pursuing an issue can approach a specific representative – allied with or independent of any political party – either because they have a local tie, or because they have an expressed interest or experience in the issue.
“Under our system, you have a choice of a) which party to go to, and b) within a party, if it had more than one representative.”
In fact, say Major and Preminger, VWR might lead to the fragmentation of existing political parties into sub-parties, and produce a quite different mode of doing politics.
“Under VWR, you’d have a cluster of parties replacing, say, the Labour Party, which would represent the different wings and dimensions of the Labour Party as it is today, and then people would attach their votes to particular sub-parties. After the election, a coalition would form depending on the strengths of the different parties and how the electorate had assigned their votes.”
If Major and Preminger are right – and I believe they are – the adoption of VWR could be just the life-saving injection our democracy needs to rejuvenate it. Their paper explores other consequences for government that we don’t have space to describe here, but they also go one step further, and propose a supplementary system which could be adopted, in time, to enhance voter representation at the heart of government: Dynamic VWR. But that’s a story for another day.
Major and Preminger’s paper is freely downloadable from JeDEM (the eJournal of eDemocracy and Open Government) (follow the pdf link at the top).
There is a shorter version on the Compass website.
The views expressed here are those of the author.