Who lost Ulsan?*

(WARNING:  There will be a bit of number crunching in this post.  Don't bother reading it if such things bore you.)

This a follow up to a post I did at the Marmot's Hole last week.  Here is the relevant part of that post:

Ulsan
The Democratic Labor Party (DLP) had expected to kept its seat in this heavily union household district, but the GNP won by 49.1% to 45.5% (hat tip to the Marmot). I could not find the numbers for other candidates. It would be interesting to know if most of the rest went to the OOP’s candidate. That loss cost the DLP their status as a floor negotiating group and they can no longer submit bills to the legislature, which might be a good thing since they are the Stark Raving Loony Party of Korean politics. If the OOP and DLP can’t learn not to run guy’s in each other’s districts, life will only get harder for both camps. (emphasis added)

My main point in that section was that, as the two main parties of the Korean left, Our Open Party (OOP) and the DLP need to avoid running candidates in each others districts or risk handing more victories to the conservative Grand National Party. 

Antti, in a comment over at Kotaji, said that the DLP was actually a victim of an OOP collapse:

In the Jinbonuri board I saw the number of votes given to each party in the last elections in that Ulsan district. The number of votes that DLP and GNP had not changed dramatically throughout the years, but the huge difference now was the downfall of OOP. It receveived only some 25% of the votes in the previous election, so much of the non-DLP votes would have moved from OOP to GNP.
GNP DLP OOP
2004년 17대총선 19900 27000 10200
2005년 재보궐선 24600 22800 2700

So it seems that DLP got the votes it was going to get with those voting percentages, but the other votes made the difference. GNP won because OOP lost so big.
The writer of the entry in Jinbonuri attributes much of this to the “Kang Jeong-koo effect”; all the noise it created should have moved conservative-minded voters towards GNP. Can’t tell, but this seems to be talked about a lot.

I certainly hope that the GNP's victory was not the result of any kind of 'Kang Jeong-koo effect' because it would encourage them try to win through cheap theatrics rather than sounder, if less exciting, policies.

I checked out the board that Antti referred to and discovered that they actually had results from the last four elections.  While every election is a unique event, comparing four should give clearer results than comparing two.  That is especially important since I contend that the 2004 election was an aberration.  Here are the (obviously rough) numbers:

Year Election-----한나라 노동당 열우당
2000년 16대총선 19000- 18500- 4000 (민주당) (general election)
2002년 지방선거 22000- 24000-- 0 (local election)
2004년 17대총선 19900- 27000- 10200 (general election)
2005년 재보궐선 24600- 22800-- 2700 (special election)

Let's see if the 2004 election was actually an aberration.  First, we need the raw vote totals.  The total votes for the major parties in each election were 41,500 in 2000, 46,000 in 2002, 58,100 in 2004 and 50,100 in 2005. 

The GNP's (한나라) percentage of the major party vote during those years were 45.8% in 2000, 47.8% in 2002, 34.3% in 2004 and 49.1% in 2005.  Their mean vote percentage for the four elections is 44.3% with a standard deviation of 5.9%.  That would indicate that three of the elections (2000,2002 and 2005) were in the normal expected GNP vote range of between 38.6% and 50.2%, while the 2004 result was unusually low.  In fact, the 2004 number is closer to two standard deviations from the mean.  That means that the low GNP total in the 2004 election could only be expected in about one out of twenty elections in the district.  The fact that the 2005 result is well within the normal range would suggest that any 'Kang Jeong-koo effect' did not give a major boost to the GNP.

Now, let's take a quick at the DLP (노동당) numbers for the same elections.  They got 44.6% in 2000, 52.2% in 2002, 46.5% in 2004 and 45.5% in 2005.  That makes a mean of 47.2% with a standard deviation of 3.0%.  All of the elections were in the normal range of between 44.2% and 50.2% except the 2002 local election when they they got a considerably higher percentage of the vote.  That was the only year in which the DLP got a one-on-one match-up with the GNP.  That would suggest to me that a the GNP was helped by the OOP's participation in the election rather than any collapse in it's support.

(UPDATE:  I guess I should point out that, in a normal distribution of four results, two of the results would be just outside one standard deviation on either side of the mean.  When only one result out of four is outside one standard deviation, it is a pretty good sign that there is something unusual about that result.)

The acid test in my book is much simpler:  First, in the four most recent elections in that district the GNP has never gotten a majority.  Second, in those elections, the DLP only got a majority when it faced the GNP one-on-one.  That reinforces in my mind that the OOP had no business running a candidate in that district unless their goal is to undercut the DLP, especially since they have never gotten more than 17.6% there.

It is possible that the OOP still hopes it can destroy the DLP and have the Korean left all to itself.  If so, then their performance in Ulsan last month worked perfectly.  But I don't think the DLP is going to fade that easily.  They will also retaliate by running people in OOP districts, which could very well cost the OOP a victory or two somewhere down the line.

In short, if the OOP and DLP can’t learn not to run guy’s in each other’s districts, life will only get harder for both camps (Yes, I have said that before).**

*(If the title of this post sounds vaguely familiar to Americans, it should be.  That link is to a red label required reading article for anyone interested in American politics.)

**(Please note that I am talking about electoral cooperation, not cooperation on policy or legislation.)

Comments

[Haksaeng]

Great post. Thanks for the additional data and for crunching the numbers.

[kushibo]

I'll have to digest all the number-crunching later, but for now I'll tell you that I got a chuckle out of this: which might be a good thing since they are the Stark Raving Loony Party of Korean politics.

Damn right!