Han recall, four weeks out

A little less that four weeks before the Kaohsiung mayoral recall vote, it isn’t looking really great for Mayor Han.

Remember, two conditions are necessary for the recall to succeed. Yes votes must outnumber no votes, and yes votes must be at least 25% of the eligible voters. The overall turnout rate does not matter.

There is a moderate amount of polling from pollsters of varying quality. However, the results are startlingly consistent. The wonderful Wikipedia editors have collected five polls since the beginning of February.

pollster date turnout yes no Turnout + yes
Apple May 8 47.7 51.3 33.0 37.3
INA May 5 47.5 55.1 32.0 36.0
NPP Apr 20 43.7 52.1 35.2 33.2
TBT Mar 20 51.2 59.5 34.5 39.5
TVBS Feb 7 44 53 32 34.8

All of them show that around 45-50% of respondents say they will vote. By about a 5 to 3 margin, people saying they will vote yes outnumber people who say they will vote no. There are some undecided voters, but with such a large margin of yes over no, it seems pretty unlikely that Han can defeat the turnout by mobilizing all his supporters to vote no.

If the first condition is likely to be satisfied, that means that Han’s best chance is for the yes votes to fall below 25% of eligible voters. These polls all show votes for the recall in the mid to high 30s, well above that threshold.

Of course, survey respondents always tell pollsters that they will turn out to vote. If memory serves me correctly, in the recent presidential election, in most polls somewhere around 90% of respondents claimed that they planned to vote. In fact, turnout was just under 75%. However, I think that turnout near 50% is not an unreasonable expectation. Lots of by-elections for legislators or even township mayors get 40% turnout, so it doesn’t seem unlikely at all that you would get 50% turnout for such a high-profile recall as this. There two reasons for this. One is that higher offices produce higher turnout. Direct municipalities are springboards to the presidency. All things equal, we should expect more interest in this race than in a legislative or township mayor race. Moreover, Han is an extremely well-known politician who arouses strong feelings.

In other words, if nothing dramatic changes between now and June 6, I think it is pretty likely that the recall will succeed.


Han has thus far tried a few things to turn back the recall. First, he has tried mobilizing his own loyal supporters to come out and vote no. He has some very loyal and ardent fans, but the polls show little promise for this strategy. Second, he has tried positioning himself as a serious and conscientious mayor who is concerned with public health. I don’t think this is working either. Attitudes about Han were baked pretty solidly during the presidential election; I doubt he can reshape his image this quickly with this little media attention. Third, he has tried to suppress turnout through his control over the city government machinery. Lots of the usual polling places have tried to refuse to be available as a polling place this time, claiming that serving as a polling place would clash with their anti-Covid responsibilities. I don’t think this is going to work either. On the one hand, the Central Election Commission is working hard to ensure that there are enough polling places available. On the other hand, voters who think that a politician is trying to deprive them of their right to vote often turn out in even higher numbers. The wider KMT is not going to actively support vote suppression tactics either. Neither party has a tradition of vote suppression, and being tarred with that label could have devastating long-term consequences.


Han’s best hope is that the Covid-19 pandemic will scare large numbers of voters into thinking that it is too dangerous to turn out. That is, actively trying to suppress the vote might not work, but passively (or through underground murmurs) hoping that people are scared might work. What would really help Han is if there were a few cases of domestic transmission over the next few weeks to terrify voters. In a nutshell, the Central Epidemic Command Center (CECC) is making Han’s recall more likely by successfully keeping Taiwan virus-free.

There is a high-stakes game going on with relaxing restrictions. Every time the government announces a relaxation – allowing more fans into baseball games, encouraging tourism, relaxing social distancing on public transportation – it sends a message to Kaohsiung voters that society is safe. The safer people feel, the more likely it is that they will come out to vote. However, relaxing restrictions simultaneously makes it more likely that any person who has the virus and is out in society will infect a higher number of people. News of such a disaster would almost certainly lower turnout and might push the number of yes votes below 25%. Of course, I don’t think the government’s primary consideration is the Kaohsiung recall, but the success or failure of that recall may nonetheless depend on the success or failure of their relaxation measures.


I love the irony of Han’s fate depending on governance. Han is a populist who claimed that the DPP government was hopelessly corrupt and therefore bad at governing. Now, his best (only?) chance of survival is if the DPP government, which has been so glaringly competent for the last year and a half, suddenly morphs into the incompetent regime of Han’s rhetoric. On the other hand, if it was an ungrounded attack all along, Han is probably doomed. President Tsai’s good governance crushed his populism in the presidential election, and now even better performance from her administration this year might be the death blow to his mayorship.


2 Responses to “Han recall, four weeks out”

  1. Red Says:

    Hi Garlic, I am a bit confused by the chart table. The math seems off, unless I’m misunderstanding it.

    In the most recent Apple poll listed there (May 8,) turnout was estimated at 47.7%, and 51.3% of voters say they would vote yes, but that only comes out to 24.47% for yes in total, which falls barely short of the 25-percent threshold. Why does the right-most column say 37.3% then?

    • frozengarlic Says:

      I copied that from the Wiki page; I didn’t do the math myself. I assume they didn’t only ask people who said they would definitely vote how they planned to vote. That is, they may have asked all the people who said they probably wouldn’t vote how they planned to vote if they did vote. Since a higher proportion of the likely non-voters are against the recall, that could account for the discrepancy. We are in the blackout period now, so I can’t check survey results. Sorry for the slow response.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: