Presidential campaign accounts

A few days ago, the Liberty Times published a story about the financial accounts for the Tsai and Han presidential campaigns. I rarely talk about the role of money in politics because we generally don’t have very good data to look at. This is real data, so maybe it’s worth a quick glance.

However, before I talk about the actual data, let’s quickly note that there are limits to these data. These data come from the Control Yuan, which collects data on political donations based on legal requirements established in various versions of the so-called “Sunshine Laws.” Unfortunately, there are lots of shady spots in the Sunshine Laws, so you can’t expect that these data are the full story. There is not really any penalty for misreporting. If someone shows evidence that a politician has misreported, the politician can simply file a new, accurate report within a certain period of time. Moreover, these reports only cover the accounts of the Tsai and Han campaigns, narrowly construed. They don’t cover all the spending of other organizations, such as the DPP and KMT, various legislative candidates who helped out, or outside supporting groups, such as the Friends of Tsai Ing-wen (小英之友). Finally and most importantly, it isn’t clear to me what the time period is. A quick glance at the Control Yuan website seems to indicate that most of these data start from Sept 1, 2019. Of course, by Sept 1, both campaigns had been in full swing for seven or eight months. So, these data are by no means the full story. Still, they are at least part of the story. And since I don’t think contemporary presidential campaigns are like, say, legislative campaigns from the early 1990s in which as much as 90% of the actual expenses were illegal (read: vote-buying) and off-the-books, these are probably not completely worthless data.


[A quick note on currency and numbers. Even after all these years, my brain still doesn’t process large numbers in NT dollars very well. This has something to do with the basic difference between English numbers, which are based on three zeroes (thousands, millions, billions, trillions) and Chinese numbers, which are based on four zeroes (ten thousands 萬, hundred millions 億, trillions 兆). I instinctively understand small numbers, since I live with those in my everyday life. After a brief experience house hunting in Taipei several years ago, numbers up to tens of millions 千萬 also make sense to me. But once you start to get in to the hundred millions 億 and above, my eyes glaze over and I need to start converting into US dollars for anything to make sense. So I’ll give you the data in both NTD and USD, but I’m going to discuss everything in terms of USD.]


So how much does it cost to run a presidential campaign? Tsai’s reported budget is roughly USD 20 million, while Han’s is roughly USD 15 million. As noted above, this doesn’t cover the entire campaign, but it probably covers most of the last four months, when the campaign expenses were probably at their highest. My very rough and uneducated guess is that the total cost for the entire campaign (starting at the beginning of 2019) was probably about three times these amounts. But I could be way off with that guess, so don’t quote me on it. That’s real money, but a hell of a lot less than American presidential candidates are burning though, even considering the difference in population size.


Liberty Times broke down both the revenues and expenses into several categories. Let’s look at donations first. In NTD, they are as follows:


Tsai Ing-wen (DPP) Revenues (NTD) Han Kuo-yu (KMT)
          338,148,598 Individuals           371,129,494
          160,569,287 For-profit businesses             45,263,389
                   14,500 Political parties               6,000,000
              6,442,000 NGOs               1,230,400
            59,565,732 Anonymous sources             32,676,370
                   21,512 Other                    11,170
          564,761,629 Total revenues           456,310,823


Divide the numbers by 30 to convert to USD, and you get the following:


Tsai Ing-wen (DPP) Revenues (USD) Han Kuo-yu (KMT)
        11,271,620 Individuals       12,370,983
          5,352,310 For-profit businesses          1,508,780
                     483 Political parties             200,000
             214,733 NGOs               41,013
          1,985,524 Anonymous sources          1,089,212
                     717 Other                     372
        18,825,388 Total revenues       15,210,361


The two candidates got roughly the same amount from individual donations, but Tsai got significantly more from private enterprises. This latter gap is both entirely predictable and simultaneously a bit surprising. On the one hand, businesses want to buy influence, so they always donate more heavily to candidates who look likely to win (and will be in position to make decisions). Since Tsai was clearly leading by September, it isn’t surprising that businesses bet on her. On the other hand, businesses have traditionally favored the KMT. This is especially true for all those businesses that have operations in China. They want traditionally want a government who will help them access the China market, and some of them have to openly support the One China candidate in order to satisfy pressures from within China. So it’s a little surprising to me that all those Taishang didn’t donate more than USD 1.5 million to the Han campaign, just a public relations ploy to their Chinese associates.

The fact that Han actually got a bit more than Tsai in individual donations is also striking, given that Han was trailing in the polls. We heard stories all year about the fervent Han fans, and this might be a reflection of that die-hard support.


What about the expenditures? Here are the data in NTD and USD. Note that I have translated 宣傳  as “messaging.” Sometimes this term gets translated as “propaganda,” which has always made it sound somewhat Soviet to me. Other possibilities, such as “advertising” or “communications” don’t quite seem to capture the broadness of this category.


Tsai Ing-wen (DPP) Expenditures (NTD) Han Kuo-yu (KMT)
          325,861,233 Messaging           381,286,374
          101,796,817 Events             23,511,551
            67,913,244 Personnel               6,388,387
              6,590,742 Transportation               3,401,994
            29,209,856 Rent and office                  932,371
            57,524,974 Other             10,655,529
          588,896,866 Total           426,176,206



Tsai Ing-wen (DPP) Expenditures (USD) Han Kuo-yu (KMT)
        10,862,041 Messaging       12,709,546
          3,393,227 Events             783,718
          2,263,775 Personnel             212,946
             219,691 Transportation             113,400
             973,662 Rent and office               31,079
          1,917,499 Other             355,184
        19,629,896 Total       14,205,874


Both campaigns spent more on messaging than any other category. However, while Tsai spent about half of her budget on messaging, Han reported spending nearly 90% of his money on this category.

The media stories were all about how the campaigns spent money to appear on new media. They had a field day breathlessly telling us that Han spent two and a half times as much as Tsai for his appearance on the Nite Nite Show. I guess I’m just not that interested by that little story. I’d like to know how much they spent on You Tube shows, Facebook, Line, traditional media, and so forth. But we don’t have that data here.

Han reports spending USD 1.5 million on everything else. This seems a bit unlikely to me. I will believe that he didn’t spend anything on renting out campaign offices, but only if you tell me that people donated space for little or no cost. Of course, that should be counted at market rates (for both donations and expenses), but this is one of those shady spots in the Sunshine Laws.

The gap in personnel costs is stunning. Tsai reports spending ten times as much as Han on staff! I would be willing to believe that Tsai ran a fully staffed and professional (read: expensive) campaign while Han had to make do with a relatively amateur operation, but this difference is astonishing. After all, the KMT has ample experience running real campaigns; they should have minimum standards for staffing.

The gap on events, where Tsai reported about four times more spending, seems more reasonable to me. We all remember Han’s big, spectacular events. However, as someone who spent the campaign looking for events to attend, it was my clear impression that the Tsai campaign put on a lot more events. A lot of them were smaller, but those all cost money. She was doing several events nearly every day, while he seemed to drop off the campaign trail for two or three days at a time. This also explains the gap in transportation expenses.


Several times over the past few years, I have thought about how the KMT and DPP seemed to have switched roles. These expenditure reports are another example. 25 years ago, the DPP would have been running the shoestring campaign with no professional staff, no fancy offices, and spending everything they could on their message. The KMT was the one with all those overhead costs. I don’t think the DPP has gotten to the old KMT levels of unproductive bureaucratic bloat, but these numbers suggest that they are now the ones running the big, slick, professional operation.

2 Responses to “Presidential campaign accounts”

  1. helder Says:

    Good report, but nothing takes me away from the idea that Han already knew he would lose the elections. How come there is around 1 million dollars difference from the revenues? If Tsai has a clearer account than Han, then isn’t the unaccounted expense somehow ambiguous? Smells tricky. Smells corruption. Smells like Han made an election run with the point of earning some money for himself. Maybe he himself is not alone on the “job”, but sure fells like he is in it.

  2. kezza Says:

    Nathan, please consider yourself lucky to be in Taiwan rather than in the Indian subcontinent, where digits are grouped by rightmost 3 and then every pair to the left (more significant). That sure confuses everybody.

    Anyway, a few points:

    (1) I am willing to believe Han didn’t spend much on rent and office. After all, we heard stories throughout 2019H2 that KMT was willing to lease the top-floor of their Taipei office building for something like 200k NTD per month but in the end Han’s camp didn’t pay a single dime of rent to the KMT. The use of Kaohsiung’s KMT office as his base was “voluntary donation” in a similar fashion. Those costs are on the KMT side of the books and probably part of the 6m party donation package (which again, isn’t marked-to-market).

    (2) Personnel cost. Again most of that cost is on KMT’s side, although we heard a lot about conflicts between KMT staffs and Han’s inner circle. KMT leases its staff to Han’s camp but whether they were actually used by Han for anything is still unclear. Han’s figure pretty much only paid for his inner circle friends (Peter Pan, Anne Wang, …) and part-time college students on minimum wage.

    (3) Businesses always hedge their bets, with only few exceptions. So while the Chinese operation had to satisfy certain minimum amount of donation to Han et al (which I don’t think will be in the books, e.g. directly funnel through TLO-affiliations instead) they have their other Taiwanese subsidiaries making comparable donation to Tsai. So Tsai-Han difference of just over 100m NTD weren’t really that surprising to me, especially if the rumours about Wu preempting Han’s fundraising within traditional KMT business circle is true.

    (4) I’d translate “宣傳” as Promotions, which in my dictionary covers a bit more than advertising. It covers pretty much everything from on-ground word-of-mouth and posters, SWAGs to “fake news” that Nazis/Soviets/… could only envy.

Leave a Reply

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

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

Facebook photo

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

Connecting to %s

%d bloggers like this: