Tuesday, December 16, 2014

Proportional Representation and Japan

While Wildrose figures out what it's future is, and Glenn Thibeault decides to bring Ontario provincial politics in the news, I thought I would take a step back and examine some other issues.

The final results from Japan are out. The map I posted earlier is correct and still stands, but we now have voting numbers.

I want to, in particular, examine how these numbers apply and might work under a Parallel system, such as exists in Japan.

First, lets examine the FPTP seats; IE, the seats where one person wins, and that person is whomever happens to take the most votes.

The LDP had a overwhelming victory here, with the results as follows:

48.1% - LDP - 233 seats
22.5% - DPJ - 38
13.3% - JCP - 1
8.1% - JIP - 11
1.5% - NKP - 9
1.8% - PFG - 2
1.0% - PLP - 2
2.9% - IND - 8

There is a quick and dirty way using math to estimate the seats won in a FPTP system using popular vote. This is by squaring the popular vote result. Lets just take the top 4 parties and do this.

LDP = 231361
DPJ = 50625
JCP = 17689
JIP = 6561

total: 306236

The LDP in this case has 75.5% of the "points" as noted above. The DPJ has 16.5%. With 295 FPTP seats, we simply multiply.

LDP = 223
DPJ = 48

The other parties will be off by larger numbers, but both these examples are within 10 seats of the final result.

This is a very quick and "dirty" way of doing such estimates that will generally work in any country; but be warned, this method can not properly estimate when parties have support that is spread out (JCP) or concentrated (JIP)

What I want to do here in particular, is compare the FPTP results to the Proportional results. Here are the Proportional results on their own:

33.11% - LDP - 68
18.33% - DPJ - 35
15.72% - JIP - 30
13.71% - NKP - 35
11.37% - JCP - 20
2.65% - PFG - 0
2.46% - SDP - 1
1.93% - PLP - 0

One thing that should be clear is the radically different result this is from the FPTP results. This is somewhat misleading though as it misses some important factors. Lets just examine the governing coalition.

FPTP - 49.6% - 231 seats
PR - 46.82% - 94 seats

You can clearly see from this what happened was many LDP voters in the constituencies, voted NKP in the PR list. Why? My main suspicion is it was simply a way to keep the LDP honest, to put a cap on the power the LDP could achieve alone.

I also want to note the change between FPTP and PR results for the Government, the DPJ, and JIP

-2.8% Government
-4.2% Democratic Party
-7.0% (combined)
+7.5% Innovation Party

It is clear to me that many right-win voters split their vote between the LDP and JIP, while opponents of the government split their vote between the JIP and DPJ.

It was, clearly, more common to cast a straight ballot; that is, for the same party in both local and regional races, but only ~55% appear to have cast a straight ballot.


Why this is important is that under a Parallel system, like that Japan has, this difference matters. Under a more standard MMP model, it would not matter who you vote for locally in terms of the final seat result of the parties. The only difference is exactly who the MPs are. Most people, to be blunt, care less that their MP is Bob or Joe and care more that it's a Liberal or a Conservative.


No comments:

Post a Comment