How and when did you decide? Part 2

The majority of voters who voted for opposition parties in the recent general election actually decided to do so quite early in the campaign; nearly half of them even before Nomination Day.

Soon after that, about 60-70 percent would have decided.

This is one slice of data coming out of an online survey conducted by Yawning Bread over the Polling Day weekend. An overview of the survey and overall responses can be seen in Part 1.

The parties with stronger branding, i.e. the Workers’ Party and Singapore Democratic Party (SDP) as well as the Singapore People’s Party (SPP), led by one of Singaporeans’ favourite politicians Chiam See Tong, had very few voters deciding at the last minute — generally only 5 percent.

There are differences among the opposition parties, but they seem quite subtle.  However, these differences show up more clearly in the answers to Question 6.

The great majority of voters who voted for an opposition party had no doubt about their decision to do so. But there is a distinct pattern. Those voting for the Workers’ Party, SDP and SPP had noticeably greater certainty than those who voted for the other parties, especially if one takes into account the “no answers”. After all, if one were certain about one’s decision, it would be easier to respond to the survey accordingly; a higher percentage of “no answers” therefore suggests some difficulty in making the decision.

Nonetheless, the overall picture is that what voter support opposition parties get tend to be hard support, not the hesitant, wavering sort. This might sound like an asset, but at the same time, it may suggest a weakness: a poor conversion rate of swing voters to their side. Swing voters are by definition soft support, characterised by relatively late decisions and some doubt along the way.

That the graphs above do not show all that much soft support suggests insufficient numbers of swing voters coming over.

Of course, it’s great if a party’s hard support is so numerous that it alone can carry it to victory. In fact, that was/is the situation enjoyed by the People’s Action Party (PAP), to the extent that it might have been the reason why its communication skills atrophied. But the election results clearly show that, Hougang and Aljunied excepted, no opposition party anywhere has hard support of such numbers that victory is within grasp. There is a need to win over swing voters.

How much hard support is there?

To get some estimate of this, I cross-referenced to another set of data, namely the vote shares of each opposition party in the constituencies that it contested in May 2011.

Take the Workers’ Party. It received 46.7 percent of the votes in the areas it contested. Of these, we know (from the top bar chart) that 70.8 percent of them decided to vote for the Workers’ Party very soon after Nomination Day. Such early deciders we can consider as hard support or the party’s “reliable support base”. 70.8 percent of 46.7 percent gives us 33.1 percent.

That is to say: the estimate is that about 33.1 percent of the total electorate can be relied upon to give their vote to the Workers’ Party if they have a chance to. This is quite encouraging, for in my opinion, bedrock support of 33 – 35 percent should be sufficient as a launch-pad for electoral victory provided a party has the skill sets to win over swing voters.

Other opposition parties had lower vote-shares at the election, and also slightly lower percentages of early deciders. Using the same calculation method, we see that the bedrock support level that the Reform Party and the Singapore Democratic Alliance can rely on is just 19 percent.


You will notice that I made a calculation for the PAP as well. However, since I consider the raw data for the PAP to be very iffy, I wouldn’t put much store on the figure I obtained — 22.7 percent. The older generation with many strong supporters of the PAP were not among the survey participants in any meaningful number. Actually, I have the feeling that most of the survey participants who reported voting for the PAP were not die-hard PAP supporters anyway; they were swing voters, who this time decided to give their vote to the PAP after some careful consideration. That probably accounts for the fact that in the top chart, so many were late-deciders.

* * * * *

Analysis of the survey results comes in several parts.

Part 1: Overview of participants.

Part 2: At what point in the campaign did voters decide on their vote choice?

Part 3: How important were selected factors in determining vote choice?

Part 4: How important were selected sources of information in determining vote choice?

Part 5: What about Tanjong Pagar GRC?

12 Responses to “How and when did you decide? Part 2”

  1. 1 Gard 12 May 2011 at 09:20

    Wonder if we would live to see the day where the make-up of the parliament will correspond to the support base: WP 33.1%, SPP 28.6%, … PAP 22.7%. (Over time, the die-hard senior citizens would have cast their final dying vote for the king… which suggest the possibility of ‘afterlife voting’ in which spiritual masters under the employ of the monarchy would channel votes from beyond. Or is it possible [for the government to amend the law] to legally cast a vote after you are dead – by means of a will and proxy?)

    Mmm. Even excluding the die-hard senior citizens, you definitely have more anti-monarchist readers in your poll, that threats of repent-and-regret do not move them; or the readers have the perception that they are voting for – to borrow the phrase somewhere – estate managers.

  2. 2 Nicetalkbutwhat? 12 May 2011 at 12:22

    You have some good analysis, I wonder if you are going to forward these analysis to opposition parties and perhaps I hope you can do a recommendation or conclusion on how they should focus on turning the tide of the battle or which areas to focus.

  3. 3 whatu1 12 May 2011 at 13:45

    Sadly, I had decided where my vote will go to since I had a chance to vote. In 1988, I was abroad for my studies and didn’t get a chance to vote. Before my present flat location, I was located in Brickworks GRC in 1991, then in Tanjong Pagah GRC from 1997 till 2001.

    So from 2006 till future, it will always be the opposing side.

  4. 4 Evariste 12 May 2011 at 14:45

    Is it illegal to do cold calling, to avoid self-selection bias?

  5. 6 wikigam 13 May 2011 at 00:49

    Honestly, i can tell you that i have decide which party i will vote for on the next coming GE at 2016 . It will be 2nd lesson for them. They will never wake up .

  6. 7 Daniel 13 May 2011 at 09:55

    Hi, this dataset is very interesting! Would you be willing to share the dataset online? I can think of a few interesting crosstabs that may be worth doing: such as by breaking down moments of doubt by age or by breaking down second choice party by first choice party.

  7. 8 T 13 May 2011 at 21:05

    Thanks Alex for your efforts in designing and evaluating a Post-Election “Post Mortem”.

    a) For the first table, a sizable proportion of voters for the PAP only decided on their voting decision on Cooling Day i.e. 30% of total PAP voters

    This suggests that Cooling Day benefits the incumbent more than the other political parties, even when considering the online community. It may not just be Cooling Day itself that favors the incumbent, but the time and space that dilutes the impact of rallies which favor the opposition parties more. For these swing voters inclined towards the PAP, this allows the euphoria for change to be subtly substituted by more pragmatic thoughts of preserving/tolerating the status quo.

    b) For the third and final table presented in this article, I noticed that the % of vote share in contested constituencies trailed the % of voters deciding on/soon after Nomination Day by 23% – 29% for the opposition parties.

    To account for the under-represented PAP sample and taking an arbitrary average value of 26%,

    The PAP’s % of voters deciding on/soon after Nomination Day would be 26+37.7 = 63.7%

    It’s estimated reliable support base would thus be about
    = (63.7 / 100) * 60.1% = 38.28%

    But how much of this reliable support base is due to fear, remains to be seen. This also brings to attention a less-appreciated aspect of swing-voter politics: a fear of opposition parties themselves that is distinct from a fear of the PAP.

    c) For the second table, about 20% of voters for the PAP seriously reconsidered their vote for the party.

    This corresponds fairly well to the 30% of PAP voters who only decided on their decision on Nomination Day.

    However, an obstacle preventing the link between the results from Q4 and Q6 lies in the significant non-response rate for Q6 i.e. an average of slightly above 20%.

    Why did about 20% of respondents not answer Q6? I believe it is because they were never really truly decided on any particular party. They could indirectly explain the insufficiency of groundwork efforts, an issue that may not just be about the impossibility of knocking on every door in a constituency.

  8. 9 Randi 13 May 2011 at 23:26

    I participated in the survey and am in the minority who had several moments of doubts when casting my vote for WP – only because I was in Aljunied GRC. I thought it would have been interesting if the “Aljunied Emotional Dilemma” was factored in as well. This is in contrast with 2006, when I also voted for WP but with absolutely no doubts. Nevertheless, a great piece of work! I’m sure many Singaporeans appreciate it 🙂

  9. 10 sporean 15 May 2011 at 13:14

    The large percentge pap voters who decided in last 24 hrs,
    showed the significance of the cooling day, to the benefit of pap.

  10. 11 Harro 16 August 2011 at 10:39

    Excellent work Alex. Would be interesting to know how you collected the data.

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 )

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: