Wednesday, February 20, 2008

Man of Many Letters

Before we get into the meat of this post, I'd like to direct everyone to this great Asian Studies Toolbar by John Noyce. It has links to Asian Studies stuff on the internet, media, journals, blogs, and a host of other stuff. Installs automatically, easy to use, and customizable.... and he has fantastic taste in Taiwan blogs.

When I was younger I aspired to be a man of letters, and now that I've reached that age where my waistline expands as my hairline recedes, I've reached that goal -- letters are my life. Today a response I made to a piece of Chinese propaganda in the Yale paper a few weeks ago made its way into a letter in the Taipei Times. Stuff that gets into any media, however semi-serious, needs to be responded to, because those responses reach people.

More fun with letters occurred a couple of weeks ago when Taipei-based David Pendery responded to my response to him. It's mostly evidence-free rambling:
Again, we can see his point, but in sum, what Turton implies is that Taiwanese voters concern themselves more about a plate of free beef noodles and a little questionable assistance with a legal problem than the education of their children, the state of their economy, the environment they live in, crime and public order, the cost of housing, infrastructure development, etc. Such a claim I will not abide by.

Ultimately, I would be more inclined to agree with Jerome Keating -- whatever his political view -- when he wrote that the KMT "is not monolithic. Contrasting viewpoints abound and power struggles continue beneath the surface."

Such a view undermines Turton's apocalyptic prediction of a "permanent majority."

In any event, I am not endorsing or criticizing any one party.

Rather, I am trying to point out that respect for the various viewpoints in this country, to say nothing of more empathetic, impartial and tolerant attitudes toward Taiwanese voters and their issues (yes, their issues) is necessary to make progress in this country.
As a matter of fact, the LDP wasn't a monolith either, but it did enjoy a 38 year grip on Japan. One quirk of this exchange: the letter he responded to was a truncated version of this longer piece which the Taipei Times had already published. The TT published two versions of the same letter!....what had happened was that I sent in the first piece but about 8 days went by and it wasn't published. I figured it was too long, so I lopped off the limbs and sent it back as a response to Pendery, and they ended up publishing both. My favorite part of Pendery's letter, unmarred by numbers or critical thought, is the part where he claims he isn't "criticizing any party." His original letter said:
Instead of apportioning blame and howling about the injustice of it all, the DPP and its supporters need to wake up to reality and rein in their worst instincts. The supercilious tone of the DPP's cheerleaders, their self-righteous declamations of exactly what anyone and everyone in Taiwan should and must think and do, and their routine denigration of one half or more of Taiwan's population have gotten utterly tiresome. These are all reasons, I think, that many people are in the process of drumming the DPP out of power.
LOL. Boy, it's a good thing Pendery isn't criticizing any one party, eh?

Finally, I also sent in a letter on the problem with the districting arrangements, which as I noted in my post on gerrymandering below, favored the KMT. It didn't make the Taipei Times' grade, so here it is, gratis:

++++++++
How the Districting Affected the Election

Although the conventional wisdom assigns the DPP defeat in the Jan 12 legislative election to the disatisfaction with Chen Shui-bian, and the "failing economy", the reality is that structural factors played a key role. Among these were the districting arrangements, which favored the KMT.

According to the law, districts must be drawn by population, and must follow city and county boundaries. These sensible, strict rules generally prevent the drawing of cross-district boundaries that result in the absurd districts so common in the US. Based on these, the Central Election Commission's original proposal alloted 8 districts to Taipei City, 12 to Taipei County, and 6 to Taoyuan. For Tainan city and county, just five districts were created.

To understand the effect of this on the vote, it is necessary to look at the voting population, not at the general population. In Tainan those five districts average 283,000 registered voters each (the national average is 230,000 voters); for Taoyuan, the same average is 225,000; for Taipei County, 235,000; for Taipei City; 251,000. Of the districts that CEC created by cutting up counties, 4 of the 5 largest by voting population are in Tainan. Using the Taoyuan average as a standard, the Tainan area should have had at least one more district. Using the Tainan average for the north, roughly speaking, Taoyuan has 1 extra seat, Taipei city 1 extra seat, and Taipei County 2 extra seats.[MT: the effect is even more pronounced because if you drop the five districts on the east coast and the islands, whose populations are unusually small, the average for the populated west coast and Ilan is about 240,000 voters per district. The average district size in both Taoyuan and Taipei County is smaller than that. Any way you measure it, it sure looks like somebody shoehorned in extra districts in the north.]

Of the ten largest districts (by voting population) created out of counties, seven are in the south: the five Tainan districts and the two in Yunlin County. There is no apparent reason for this (though of course the "right" size is debateable) -- Pingtung County, with a general population of 834,000 (631,000 voters), was chopped into three small districts, while Tainan county, with a general population of 1.1 million people (847,000 voters), was awarded three giant seats. Conversely, the CEC had no trouble creating districts in Taichung County, Pingtung, Kaohsiung City, Nantou, and Miaoli of under 200,000 voters -- a figure that would have awarded Tainan at least two more seats. In Kaohsiung, where the KMT has a powerful local presence, districts averaged only 229,000 voters each.

In sum, the CEC's original proposal created smaller districts in Blue areas and very large ones in Tainan, essentially denying more than 200,000 voters in Tainan representation that was granted to voters in other districts. Size matters greatly; the KMT won 18 of the 20 smallest districts. Had the election been more closely contested, the "extra" Blue seats would have had a strong impact.

Another effect of the districting plan is the large number of safe Blue seats. Counting the islands (3), the east coast (2), Keelung (1), Hsinchu County (1), Taichung County 2 (Yen Ching-piao), the North (26), and Blue-leaning Miaoli (2) and Nantou (2), and assuming a loss here and there, almost half the seats are safely Blue. The Blues need only win half the remainder to win an overwhelming legislative majority.
+++++++++

I deleted the remainder, mostly criticisms of the DPP's election planning.

7 comments:

Tommy said...

One thing that bugged me about his letter was his claim that the KMT is not monolithic. He is indeed correct. But no political party is monolithic. The problem isn't one of monoliths. The problem is that, in Taiwan, legislators suffer consequences if they vote, or even voice opinions, that is different from the party line. You are basically elected to provide a vote for the party, which completely undermines the concept of representative democracy. Granted, even legislators in the US often don't vote exactly as their constituents would want (ideally they are supposed to). But that possibility is there.

So no, the parties of Taiwan are not monolithic. BUT they do vote in a monolithic way.

Michael Turton said...

Excellent observation.

Anonymous said...

Dear Mr. Turton,

While I find many of your observations insightful, I have to disagree on the issue of gerrymandering. Here's why:

First, it's important to note that districts were apportioned according to population, and not voting population. This might advantage Blue areas that have more expats, but it's the way it's done in many Western democracies, like the U.S. and Canada.

You can then obtain the number of seats for each city/county using the following simple (and fair, I'd say) method: take the population of each city/county as listed in Wikipedia, divide by 320,000, and round to the nearest integer (unless the result is less than 0.5, in which case you round to 1). The one county that was shortchanged was Hsinchu county, which should have had 2 seats. But of course, the current Wikipedia population figures are for Dec. 2007, well after the districts were apportioned, so some discrepancy is expected. Besides, that would have been an extra blue seat! (Also, the results in Tainan City were so close that unless the third seat were drawn to almost exactly reflect the DPP/KMT strength in the city as a whole, it would have been blue as well.)

Another way to evaluate whether the election map was fair is to ask: what would the distribution of seats have been if results from the 2004 Presidential election were reproduced? Since the results of that election were so close, a fair system would give the Pan-Green and the Pan-Blue camp a similar number of seats. In fact, transposing the results from that election gives 40 Green seats and 33 Blue seats. Along with 6 supposedly Blue Aboriginal seats and a 17/17 proportional apportionment, that gives 57 seats for the Pan-Green coalition and 56 for the Pan-Blue. Couldn't be fairer if you ask me.

It's well known that first-past-the-post systems tend to disproportionately favor the winning party, and that's exactly what happened here. Gerrymandering was not a major factor is this election.

Michael Turton said...

First, it's important to note that districts were apportioned according to population, and not voting population. This might advantage Blue areas that have more expats, but it's the way it's done in many Western democracies, like the U.S. and Canada.

Aaargh! Two things (1) apportionment by population is how the districts are justified to outsiders. No one diddling the districts will use that number. Therefore the test of whether things are fraudulent isn't to look at the general population but the voting population -- if you want to detect fraud, you need to think like a fraudster, and not take things at face value. (2) whether you use general population or voting population, the arguments are the same (do the math yourself). Just look at Yunlin, where there are two giant districts and Pingtung, where there are three districts -- but Pingtung has far fewer people than Yunlin. Nothing justifies that. Using general population makes Tainan even more striking!

The one county that was shortchanged was Hsinchu county, which should have had 2 seats.

Yes, I believe i noted that. Also shortchanged were Yunlin and Tainan. There's no way around it....there are at least 2 and probably four seats missing down south, and at least 2 and as many as 4 seats added up north. Do the math; compare Taoyuan to Tainan. There's no justification for six tiny districts in Taoyuan and 5 gigantic ones in Tainan.

320,000,

Why did you divide by this number? That is not how the seats were constructed. It has nothing to do with the reality of district apportionment.

Again -- thinking like a fraudster -- no one seriously Blue would have gambled that Tainan would produce a Blue seat. Instead, the safest thing was to do what was actually done: reduce the overall number of seats compared to elsewhere.

Gerrymandering was not a major factor is this election.

As I said, about 50 times. In a close election it would have mattered.

Anonymous said...

Response to a few points:

1. Apportionment by population is how the districts are justified to outsiders.

- You can certainly argue that apportioning by voting population is better. But by the same token you'd have to argue that apportionment in the U.S., Canada, and many European countries is equally unfair. In fact, if it were done by voting population, someone like you on the other side could say, "The DPP did it this way, which is different from how it's done in most other democracies, just to gain seats."

2. Pingtung has far fewer people than Yunlin

- Huh?! Pingtung has about 889,000 people, while Yunlin has about 725,000 people.

3. 320,000

- Whatever quota was used was probably to make the total number of single-member constituencies 73. You can argue that the KMT pushed 73 just so that the quota would disadvantage the DPP. But I think the DPP would have pushed for a different number if that were the case.

In fact, look at where the cutoffs for number of seats (if you rank jurisdictions by population) lie:

*5/6: between Taichung Co. and Taoyuan
*4/5: Changhua and Kaoshiung City
*3/4: Tainan Co. and Kaoshiung Co.
*2/3: Tainan City and Pingtung
*1/2: Hsinchu Co. and Nantou

So 3 Southern areas vs. 2 Northern/Central areas were right above the threshold and got a bit above what they deserved. 3 Northern/Central areas vs. 2 Southern areas (yes, both Tainans) were right below.

Yet another way to look at this: pick the nice round number of 300,000 as quota, and take the Dec. 2007 population figures. That would have given:

+1 to Taipei Co.
+1 to Taipei City
+1 to Tainan Co.
+1 (barely) to Taichung City
+1 to Tainan City
+1 to Hsinchu
+1 to Yilan

That's 4 more seats in Blue areas and 3 more in Green areas. Even if you remove Taichung and add Yunlin (which, at a quotient of 2.42, should actually not gain a third seat), it's 3 vs. 4, or a whopping net difference of ONE seat.

If you pick 310,000, only Tainan Co. and Hsinchu Co. would have had an extra seat.

If you're still not convinced, try to come up with an apportionment scheme that would be fair to you, and post the scheme and the seat numbers. I don't think there's much you can do to help the DPP beyond merging Kinmen, Matsu and Penghu into a single seat... (and that's still just -2 for the pan-Blue) You're welcome to prove me wrong, of course, but it'll take more than stating over and over that the KMT often commits fraud (sure, but not to the point) and repeating Tainan, Yunlin, Tainan, Yunlin, Tainan, Yunlin,...

Michael Turton said...

Hang on! I'll be back with a response.

Michael Turton said...

BTW -- I;ve NEVER NEVER said the KMT committed this fraud.

NEVER.

Michael