I was riding home last night after Chaon & Co. creamed my son and I in Axis and Allies, and got to thinking about the legislative election again (imagine that, if you can!). This time I decided to figure out how many votes the DPP would have needed to make a substantial increase in the number of seats won.

Now, obviously, if the DPP total increases by 400,000, those votes won’t be spread evenly throughout the island. The growth would be uneven. To weight the numbers, I assumed that any growth in the vote would reflect the extant situation in the district, and thus be proportional to it. For example, if a DPP candidate in District X received 10% of the entire DPP vote, then that same candidate would receive 10% of any increased vote. Naturally this is only a rough assumption, but it is a guide to what-might-have-been.

To do this I first downloaded the vote totals for each district from the CEC database. The overall constituent vote totals are here, with the DPP receiving 3,765,222 votes, and the KMT, 5,209,237. Those represent total votes for all candidates, not votes for the party. I then eliminated all other candidates, leaving only the KMT and the DPP, in most cases (in one or two cases there was no DPP candidate, or the DPP opponent was from the NPSU).

In the chart above, the first three columns are pretty self-explanatory. The fourth column represents the ratio of the DPP vote in that district to the entire DPP constituency vote. For example, in the first line, in Kaohsiung City 1, the DPP obtained 65,266 votes, representing .0173 of the entire DPP vote (1.73%). I took the first four digits, without rounding.

The next column contains two numbers. The first number is the difference between the KMT and DPP vote (KMT minus DPP). The second number is the size of the overall DPP vote increase necessary to generate enough votes in that district for the DPP candidate to overcome the KMT candidate. Again, in Kaohsiung City 1, the DPP candidate received 1.73% of the DPP vote. To generate enough votes to overcome a 27,151 vote deficit, the DPP would have had to poll another 1.6 million votes (1600000 x .0173 = 27,680). No doubt there are some errors, I probably should not be doing math after midnight with a couple of beers in me.

Of course, this also assumes that the KMT vote stays constant, which is obviously not the case. Had the DPP polled more voters, the KMT would have been out trying to stimulate even more people to come out. But I’m assuming away that effect, so we can try and get a glimpse of what-might-have-been.

For simplicity’s sake, I incremented the vote totals by 100,000. I also did not look at areas where the DPP did not run a candidate, and I ignored Jinmen and Matsu since there is no way the DPP will ever win there. To understand this, looking at the first line, if the DPP picks up an additional 200,000 votes, it will pick up the seat in Kaohsiung County 4. If it increases 400,000 votes, it gains a total of three seats (Kaohsiung City 4, Kaohsiung County 4, and Taipei County 4).

200,000
Kaohsiung County 4

400,000
Kaohsiung City 4
Taipei County 4

500,000
Taipei County 5
Changhua County 4

600,000
Kaohsiung City 3
Taipei City 2
Ilan
Taichung County 1

700,000
Kaohsiung County 1
Chiayi City

900,000
Taoyuan County 2
Taichung County 3

1,100,000
Yunlin County 2
Penghu
Taichung City 3

1,200,000
Changhua County 3
Pingtung County 2

1,300,000
Taipei County 6
Taipei County 7
Yunlin County 1

1,400,000
Taipei County 1
Taichung County 5
Chiayi County 1

1,500,000
Miaoli County 1

1,600,000
Kaohsiung City 1
Taipei City 5
Nantou County 2

1,800,000
Taipei County 1

1,900,000
Taipei County 8

2,000,000
Taipei County 10

2,200,000
Taipei City 1
Taipei City 3
Taichung City 1

2,300,000
Hsinchu City
Taichung City 2

2,500,000
Changhua County 2

2,600,000
Taoyuan County 1
Taoyuan County 4

2,900,000
Taipei City 4
Taoyuan County 3

3,900,000
Taoyuan County 6

4,000,000
Taipei City 6
Taichung County 2
Nantou County 1

4,100,000
Taipei City 7
Taoyuan County 5

5,500,000
Taipei County 11

4,400,000
Changhua County 1

5,000,000
HUalien

5,300,000
Keelung

6,600,000
Taipei City 8

8,000,000
Miaoli County 2

Based on these figures, the DPP would have had to pick up another 700,000 votes or so to gain an additional ten seats and reach a total of 40 (with the proportional at-large seats). That would have meant a roughly 20% increase in the vote total, to 4.4 million DPP votes, a figure it has never gotten within shouting distance of. If the DPP had picked up another 1.6 million votes (to 5.3 million votes), a nearly 50% increase, then it would have picked up another 28 or 29 seats.

What was the effect of the TSU? In 2004 the TSU took 750,000 votes, this time around, according to the CEC, just 93,000 people voted for the TSU, concentrated in just three areas. The DPP was unable to grab the missing TSU voters (or perhaps it did and core DPP vote plummeted, without solid survey work, there’s no way to know). Even worse, in one case, Chiayi City, the TSU candidate took 14,000 votes, while the DPP candidate lost by 7,000 (in no other district did I notice the TSU having a serious impact). It’s often assumed that TSU voters would switch to the DPP in the absence of a TSU candidate, but recall that the TSU was brought out of the KMT by Lee Teng-hui, and some of those voters may have returned to their Blue roots this election. Others couldn’t be bothered to come out (Lazy? Angered?). Clearly, the DPP’s failure to smoothly absorb the TSU and collect all its votes, the way the KMT eliminated the PFP and incorporated all its voters, potentially cost the DPP. However, in 2004 the TSU took 756,000 votes (CEC), but the geographical distribution is telling: 390,491 of those votes were concentrated in Taipei City, Taipei County, Kaohsiung City, Kaohsiung County, and Taoyuan — mostly places where the DPP was not competitive, or where it didn’t need votes. On the other hand, the DPP lost both Yunlin seats by a total of 38,000 votes — and the TSU had 33,000 votes there in 2004 (no votes in 2008).

RE the young: I just wrote this to a friend:

It’s funny because I was pretty long-term optimistic too, but I made an error, the famous DUNE error — remember, when Duke Leto errs in trusting, because he thought anyone who hated Harkonnens would never betray him? Heh. That’s what I did. I assumed that the rising Taiwan identity among the young would make them shift to the DPP over time, but I suspect it has had the opposite effect — because they are pro-Taiwan they feel safe voting for KMT candidates.

Enjoy. Discuss. Dissect.

UPDATE: The Taipei Times published my thumbnail analysis of the election.