Thread by @HatterDan, There's a lot of talk on twitter about our electoral system at [...]

There's a lot of talk on twitter about our electoral system at the moment and how representative it is of people's votes. This got me thinking what would the results have looked like under different systems,

And me being a nerd with too much time on my hands did just that

Here's an overview of the votes revived by each party in the last election

SF—535595 (24.5%)
FF—484320 (22.2%)
FG—455584 (20.9%)
Ind—266353 (12.2%)
GP—155700 (7.1%)
LP—95588 (4.4%)
SD—63404 (2.9%)
SPBP—57420 (2.6%)
Aon—41575 (1.9%)
I4C—8421 (.4%)

Source— http://irelandelection.com

Under our current system of Proportional Representation—Single Transferable Vote (PRSTV) it produced the following seats;

FF—37 (23.3%)
SF—37 (23.3%)
FG—35 (22%)
Ind—19 (11.9%)
GP—12 (7.5%)
LP—6 (3.8%)
SD—6 (3.8%)
SPBP—5 (3.1%)
Aon—1 (.6%)
I4C—1 (.6%)

Under PRSTV parties befitted as such

Over-Represented Parties
FF +1.1%
FG +1.1%
SD +0.9%
SPBP +0.5%
GP +0.4%
I4C +0.2%

Under-Represented Parties
Ind -0.3%
LP -0.6%
SF -1.2%
Aon -1.3%

(Note that I'm not considering parties who received <1% FPV or have no TDs)

Another common system to assign seats proportionately is the D'Hondt method. This system is used for assigning cabinet positions in the Northern Ireland Executive and was use in European elections in Great Britain (NI used PRSTV)

D'Hondt would have given the following results.

Numbers in square brackets show the difference compared to PRSTV

SF—40 (25.2%) [+3]
FF—36 (22.6%) [-1]
FG—34 (21.4%) [-1]
Ind—20 (12.6%) [+1]
GP—11 (6.9%) [-1]
LP—7 (4.4%) [+1]
SD—4 (2.5%) [-2]
SPBP—4 (2.5%) [-1]
Aon—3 (1.9%) [+2]

Under D'Hondt, parties would've befitted as such

Over-represented Parties
SF +0.7%
FG +0.5%
FF +0.4%
Ind +0.4%

No Discrepancy
LP +0.0%
Aon +0.0%

Under-represented Parties
SPBP -0.1%
GP -0.2%
SD -0.4%

The issue with this model is that D'Hondt uses a list system for voting. As such it is much harder for individual independents to get elected under the system. Of the 70 former GB seats in the European Parliament not one independent MEP was elected.

This is what the Dáil would look like under D'Hondt with only party lists and no independent lists

SF—46 (28.9%) [+9]
FF—41 (25.8%) [+4]
FG—39 (24.5%) [+4]
GP—13 (8.2%) [+1]
LP—8 (5%) [+2]
SD—5 (3.1%) [-1]
SPBP—4 (2.5%) [-1]
Aon—3 (1.9%) [+2]

Under party-list only D'Hondt, parties would have befitted as such

Over-represented Parties
SF +4.4%
FF +3.6%
FG +3.6%
GP +1.1%
LP +0.6%
SD +0.2%

No Discrepancy
Aon +0.0%

Under-represented Parties
SPBP -0.1%

Within the Dáil, Independent TDs and TDs from small parties form Technical Groups. They are currently three such groups;

•Regional Group (8 Ind +1 Aon)
•Rural Group (6 Ind)
•Independent Group (5 Ind + 1 I4C)

Under D'Hondt, independent candidates sometimes from independent lists to contest the election under

In this scenario I have divided the independent vote proportionably to the number of TDs in each group

What the Dáil would look like under D'Hondt with independent lists

SF—40 (25.2%) [+3]
FF—37 (23.3%) [+0]
FG—34 (21.4%) [-1]
GP—11 (6.9%) [-1]
Regional—8 (5%) [+0]
LP—7 (4.4%) [+1]
Rural—6 (3.8%) [+0]
Ind—5 (3.1%) [+0]
SD—4 (2.5%) [-2]
SPBP—4 (2.5%) [-1]
Aon—3 (1.9%) [+2]

Under D'Hondt with independent lists, parties would have befitted as such

Over-represented Parties
SF +0.7%
FG +0.5%
FF +0.4%

No Discrepancy
LP +0.0%
Aon +0.0%

Under-represented Parties
SPBP -0.1%
Ind -0.2%
GP -0.2%
SD -0.4%

Why do we care about all this fancy proportional representation? Many democracies such the USA and UK use First Past The Post (FPTP)

In FPTP whoever gets the most votes in a constituency, (no matter how small that vote is) takes the seat. How would this apply to the Dáil?

I made a few assumptions about how to make this work. FPTP uses single seat constituencies. I considered looking at tallies & creating single seats from this data, but there would be a wide variance depending on how I drew the boarders, as such I used our current constituencies

I took whatever party topped the poll in each constituency and gave them seats depending on the size of the constituency.

Source - Wikimedia Commons

With this method of estimating FPTP results, the Dáil would look like

SF—63 (39.6%) [+26]
FF—46 (28.9%) [+9]
FG—29 (18.2%) [-6]
Ind—21 (13.2%) [+2]

Under this estimation of FPTV, parties would have benefited as such

Over-represented Parties
SF +15.1%
FF +6.7%
Ind +1%

Under-represented Parties
FG -2.7%

Parties Without Representation
GP, LP, SD, SPBP, Aon

One of the issues with this estimation is that most countries that use FPTP drift towards a two party system (Democrats and Republicans in the US and Labour and Conservatives in the UK).

So what would it have looked like if we had an FPTP election along left vs right lines?

To get this estimation I looked at each constituency and what share of the vote all candidates left of centre got vs candidates right of centre.

Whoever got more I gave the constituency to.

This estimate of FPTP with a two party system gives us

Right—94 (59.1%) [+7]
Left—65 (40.9%) [-7]

The difference from how people voted would have been

Right +4.2%
Left -4.2%

As was rightly pointed out, it is very rare for the D'Hondt method to be used on a national level. Most often it is applied on the level of constituency

To do this I kept all constituencies the same size and used D'Hondt to chose who gets the seats. I am using independent list

D'Hondt on a constituency level would produce this Dáil

SF—50 (31.4%) [+13]
FF—42 (26.4%) [+5]
FG—41 (25.8%) [+6]
Ind—17 (10.7%) [-2]
GP—4 (2.5%) [-8]
SD—2 (1.3%) [-4]
LP—1 (.6%) [-5]
SPBP—1 (.6%) [-5]
Aon—1 (.6%) [+0]

Under Constituency Level D'Hondt, parties would've befitted as such

Over-represented Parties
SF +6.9%
FG +4.9%
FF +4.2%

Under-represented Parties
Aon -1.3%
Ind -1.5%
SD -1.6%
SPBP -2%
LP -3.8%
GP -4.6%

So quite a few members of Fianna Fáil have questioned me categorising them as a right of centre party in the FPTP scenario https://twitter.com/malcolmbyrne/status/1273035113677324288

https://twitter.com/malcolmbyrne/status/1273035113677324288

I've put together two more FPTP scenarios that include a third party, similar to how the Liberal Democrats effect elections in the UK

The first is where FF, centrist inds & centre left parties from a centrist party

The second is where FF & centrist inds form a centrist party

The first estimation using a centrist party made from FF, centrist inds & centre left parties gives us

Centre—75 (47.2%) [+11]
Left—43 (27%) [-2]
Right—41 (25.8%) [-9]

The difference from how people voted would have been

Centre +8.7%
Left -1.9%
Right -6.8%

The second estimation having only FF & centrist inds comprising the centrist party gives us

Left—107 (67.3%) [+38]
Right—45 (28.3%) [-5]
Centre—7 (4.4%) [-33]

The difference from how people voted would have been

Left +24.4
Right -4.7%
Centre -19.7%

Latest Threads Unrolled: