Reasons to be Impossible

So, I was at the count listening to the BNP counting agent recounting why he thought he "mixed race" grandson he was raising should be entitled to join the BNP, because he is culturally British - though, he said, he understood it is important to protect the species.

I really didn't want to kick off, it was a count, and I was surrounded by them, so I let it go.

Their election agent later explained the election strategy to me - they weren't standing any constituency candidates, because if Labour and Conservatives got more of those, they';d be entitled to fewer list seats.

Clearly, they don't understand the additional member system and the d'Hondt rules.

In the GLA there are 25 seats. 14 are elected in single member constituencies under First Past the Post. 11 come from the list. As you can see, there are more FPTP seats than list, so the first thing to note is that a party can win more direct seats than their proportion of votes would allow. This is especially the case considering that plurality elections allow the biggest minority to win, so, say, a vote of 26-35% could be enough to take each seat.

Labour and Tories did split the 14 seats between them, 8 and 6 with 37% & 28% respectively. (that is 32% and 24% of the seats - a subproportional tally).

The way the d'Hondt works, is you take the total number of votes cast for the list, and divide it by one plus the number of seats gained.

Party	Con	Lab	Lib	Gre	BNP
# Votes	835,535	665,443	252,556	203,465	130,714
% Votes	34.05%	27.12%	11.22%	8.29%	5.3%
# Seats	11	8	3	2	1
% Seats	44%	32%	12%	8%	4%

As can be seen from this table, the parties who only received their d'Hondt votes basically got close to their proportional vote in representation (actually, the BNP are under represented, but, then, the whole problem here is one of rounding, and the rounding is only 1% in this instance.

Here's how the seats would have looked if the election had only been a d'Hondt list based one:

Party	Con	Lab	Lib	Gre	BNP
# Votes	835,535	665,443	252,556	203,465	130,714
% Votes	34.05%	27.12%	11.22%	8.29%	5.3%
# Seats	11	8	3	2	1
% Seats	44%	32%	12%	8%	4%

Look familiar? d'Hondt actually favours larger parties anyway. If we exclude the 15% of votes for parties that failed to get representation (i.e. only take our percentages from effective votes) we get the following table:

Party	Con	Lab	Lib	Gre	BNP
# Votes	835,535	665,443	252,556	203,465	130,714
% Votes	40.02%	31.87%	12.10%	9.75%	6.26%
# Seats	11	8	3	2	1
% Seats	44%	32%	12%	8%	4%

Which is, in effect, the percentages if these had been the only parties standing and voted for.

Now, you can manipulate the additional member system, if you stand as two parties, or, if you have a sister party standing in the FPTP seats you can swing a few constituencies to. But the BNP thinking was bunkum, they neither gained nor lost from not standing in any FPTP seats.

Just proof of stupidity.

Labels: Additional Member System, BNP, d'Hondt, Electoral systems, GLA Elections, Voting