Tag Archives: proportional representation

Getting things in proportion

4 Jun

PR Council

Newham council would look very different under PR

Last week the Hackney Citizen reported that in their borough the Greens would have won seven seats under a proportional voting system, whereas they actually got none – despite getting more than 20% of the vote.  In total PR would have delivered twelve opposition councillors, rather than the seven actually elected (four Tories, three LibDems).

Things in Newham are even worse.

For the second election running, not a single opposition councillor was elected, despite one-third of voters choosing candidates from parties other than Labour. According to the council’s own results webpage, the Conservatives got 24% of the vote.

Using exactly the same analysis as the Hackney Citizen – applying the proportional representation system used in the European elections to the Newham results – the Tories would have won twelve seats, making them the largest opposition group. And, despite standing in only a handful of wards, UKIP would have two councillors.

Labour would have won all three seats in six of the 20 wards – Boleyn, Canning Town North, Forest Gate North, Little Ilford, Stratford & New Town, and West Ham – and two in all the rest. UKIP would have taken a seat in Canning Town South and in Custom House, with the Tories taking a seat in each of the 12 remaining wards.

How did I work this out?

The system used in the European election relies on voters selecting a single party, rather than vote for individual candidates as they do in ‘first past the post’ elections. To get around this I averaged the votes for the candidates for each party. I did this in every ward.  From this I could then apply the D’Hondt method to calculate the results.

Taking Forest Gate South as an example, this is the actual result:

Candidate Party Votes
Masihullah Patel Labour 2209 Elected
Dianne Walls Labour 2095 Elected
Winston Vaughan Labour 2023 Elected
Mahboob Rizu Ahmed Conservative 993
Asif Choudhary Conservative 976
Tim Roll-Pickering Conservative 693
William Heron Liberal Democrat 293
Niall Mulholland TUSC 238
Dieutane Jean Parson Christian Peoples Alliance 179
Malcolm Williamson Christian Peoples Alliance 159
Ionel Vrancianu Independent 101

This gives an average party vote of:

Party Votes
Labour 2109
Conservative 887
Liberal Democrats 293
TUSC 238
CPA 169
Independent 101

Since Labour has the most votes they win the first of the three seats available.

For the second seat the votes for each party are divided by the number of seats they have won, plus 1. So Labour’s vote is divided by two and the other parties are divided by one (so they stay the same), which gives us:

Party Votes
Labour 1055
Conservative 887
Liberal Democrats 293
TUSC 238
CPA 169
Independent 101

Labour still has the most votes and wins the second seat.

For the third and final seat the votes for each party are again divided by the number of seats they have won, plus 1. So now we divide the original Labour vote by three, as it has already won two seats, and the other parties stay the same:

Party Votes
Labour 703
Conservative 887
Liberal Democrats 293
TUSC 238
CPA 169
Independent 101

In this round the Conservatives have the most votes, so they win the last seat. Forest Gate South has two Labour and one Conservative councillor.

A note of caution

These are only approximations since there’s no way of calculating with absolute certainty the result under proportional representation because of the differences between this system and first past the post. There’s also no way of knowing if changing the voting system changes people’s voting behaviour. Or if a different voting system would encourage more parties to stand candidates in more seats.

But it does provide a reasonable basis – using real results and a system actually used in the UK – to argue that our current voting system for local government is broken. It delivers an unfair result and a council that does not truly reflect the diverse political opinions of the community it is meant to serve.

Advertisements