Proportional Influence
At the time of writing, the results declared for yesterday’s election are as follows:
| Party | Seats | Votes | Votes Per Seat | Weight |
|---|---|---|---|---|
| Democratic Unionist Party | 8 | 168,216 | 21,027 | 0.49 |
| Labour | 251 | 8,335,311 | 33,208 | 0.78 |
| Conservative | 294 | 10,296,598 | 35,022 | 0.82 |
| Social Democratic & Labour Party | 3 | 107,396 | 35,799 | 0.84 |
| Sinn Fein | 4 | 150,638 | 37,660 | 0.88 |
| Alliance Party | 1 | 42,325 | 42,325 | 0.99 |
| Plaid Cymru | 3 | 165,394 | 55,131 | 1.29 |
| Scottish National Party | 6 | 482,823 | 80,471 | 1.89 |
| Liberal Democrat | 52 | 6,518,970 | 125,365 | 2.94 |
| Green | 1 | 271,294 | 271,294 | 6.37 |
| Overall | 623 | 26,538,965 | 42,599 | 1.00 |
Obviously, it’s well-known that our first-past-the-post electoral system often produces highly non-proportional results, and this illustrates this: whereas it took only about 21,000 votes to elect each DUP member (on average) it took over 270,000 votes to elect Caroline Lucas, Brighton’s new Green MP.
There are many alternatives to first-past-the-post, each having with different degrees of “proportionality” and strengths and weaknesses. Ignoring people who don’t even in principle support proportionality, the strengths and weaknesses of various systems generally revolve around the following points
- Many people dislike list systems (particularly closed lists) because voters lose the ability to favour particular candidates (arguing against pure list systems and AMS)
- Many people feel the geographic link is important, and with those, many also prefer single-member constituencies (arguing against lists and STV)
- Some people dislike voters being asked to do more than mark a single cross by a single candidate (arguing against STV, AMS, AV)
- Some people dislike hybrid systems like AMS because they create “two classes” of MPs (geographically elected MPs and top-up/list MPs).
- Some people worry about tiny parties with little support gaining seats (arguing for some de minimis, while feeling this should not be arbitrary).
All of the electoral systems I know focus on achieving proportionality of influence (by party) by allocating seats proportionately.
It occurred to me today (as, no doubt, it has to others before) that a a more direct alternative would be to weight the influence of MPs so that proportionality of influence within parliament is achieved without needing to change the voting mechanism from the present one at all.
The last column in the table of results shows the weight that would be given to MPs of different parties under this variation, on the assumption that all MPs from a given party would be weighted equally. So here, Lib Dems MPs would each get votes weighted at 2.94 whereas Labour MPs would get votes weighted at 0.78.
The system could be refined further by making the influence of MPs within their own party proportional to their constiuency success, as measured either by the number of votes they receive, or the proportion of the vote they achieve. Given the different constituency sizes, the latter is probably better.
With this refinement, if one Labour MP got 50% of the vote in his/her constituency and another got only 25%, the first MP’s vote would be worth twice as much as that of the second.
The obvious “disadvantage” with this scheme is that it adds some complexity to voting within parliament. But this seems like a tiny disadvantage compared with its advantages (to me).