DPP votes in Aboriginal townships, part 2

In a recent post, I pointed out that the DPP received an unprecedentedly high vote share in the 30 predominantly Aboriginal townships. This seems to be evidence that preferences are changing among Aboriginal voters. Even if those votes reflect opposition to the KMT rather than support for the DPP, at least the DPP party label is no longer the ballot box poison that it seemed to be in the past.

In a comment on another post, Joseph Wang offered an alternative hypothesis. Maybe large numbers of Han people are moving into these traditionally Aboriginal townships. The DPP’s higher popularity among the increasingly populous Han residents might be what is driving the overall rise in DPP vote share. This seemed quite a reasonable suggestion to me, so I thought I’d look into it.

There is no way to tell Aboriginal and Han voters apart in the executive elections I looked at in the previous post. However, in legislative elections Aborigines have separate districts, so we can count how many Han and how many Aboriginal voters there are in each township. I looked at the 30 townships with a predominance of “mountain Aborigines” (roughly speaking, all tribes except Amis). This time, I also looked at 5 townships in which at least half of the population was “plains Aborigines” (ie: Amis).

	30 mountain	5 plains
Eligible Aboriginal Voters (1992)	93851	27764
Eligible Han Voters (1992)	31921	27162
Eligible Aboriginal Voters (2012)	115473	26066
Eligible Han Voters (2012)	32490	21683
% Aborigine (1992)	74.6	50.5
% Aborigine (2012)	78.0	54.6

It doesn’t seem to be the case that Han migration is driving the trends. The Aboriginal townships have a higher overall percentage of Aboriginal voters in 2012 than they did in 1992. It’s also not the case that individual townships. The biggest drop was in Lanyu 蘭嶼 (8.0%), followed by Namaxia 那瑪夏 (4.1%), Wulai 烏來 (3.8%), Chenggong 成功 (0.5%), and Wutai 霧臺 (0.1%). The other 30 townships all saw the percentage of Aboriginal voters increase.

If the number of Han voters is not increasing, it is certainly possible that their political preferences have changed over the past 20 years (just as in the rest of Taiwan). I have heard one story about Han people in Aboriginal townships countless times over the years. Many (most?) of them are Mainlanders, retired soldiers who married an Aboriginal woman and moved into the Aboriginal community. I don’t know how much truth there is to this stereotype. At any rate, it is possible that the Han population in Aboriginal townships has evolved, either in its demographic composition or simply in its political preferences.

Has the DPP gotten more votes among Han voters in Aboriginal townships over the past 20 years? Again, this can be examined in legislative elections, where Han voters choose from a different set of candidates than Aboriginal voters. The 30 mountain townships are probably a better indicator since they encompass a wide range of electoral districts and are thus far less sensitive to variations in candidate quality. (I don’t think it is advisable to duplicate this methodology among Aboriginal voters since the quality of DPP candidates in those elections varied too widely.)

	30 mountain	5 plains
DPP votes (1992)	3416	3281
DPP votes (2012)	7063	4130
DPP vote share (1992)	14.9	19.1
DPP vote share (2012)	25.7	24.8

The DPP has clearly made inroads among Han voters in Aboriginal townships. In the 30 mountain townships, the DPP’s vote share was about 10% higher in 2012 than twenty years earlier. However, remember that only about a quarter of the electorate in those areas was Han. That implies that, if there was no change among Aboriginal voters, the DPP vote share in executive elections should have gone up by only about 2-3% of the past 20 years. The DPP’s increases prior to this year were modest, but they seemed to be more on the order of 5% or so. Han voting might be one part of the changes, but there had to be some changes in Aboriginal voting behavior as well.

All this says very little about the DPP’s huge spike in 2014. That change is so large and so sudden that it simply cannot be explained by changes among Han voters. Aboriginal voters had to have changed as well.