On Politics
Published in

On Politics

2020 Sri Lanka General Elections

The Split Effect

A Hypothetical Experiment

A friend of mine claimed that the UNP + SJB split would actually be beneficial. And that the split would win more seats, than if they contested the 2020 Sri Lankan General Election under a single party. My friend’s reasoning was something like this: “Many would-be UNP-voters would not vote UNP, because Ranil is leader. Similarly, many would not vote UNP because of Sajith. With the split ticket, the Anti-Rs would vote SJB. And the Anti-Ss would vote UNP.”

Source: https://colombogazette.com/2019/09/10/ranil-and-sajith-hold-do-or-die-talks/

Many in my Twitter audience (a highly biased and statistically insignificant sample) seem to agree.

There might be truth to these beliefs. However, there is another factor that might put a spanner-in-the-works: Our (not so) proportional representation system itself.

In (Not so) Proportional Representation, I discussed how our proportional representation system is “not completely proportional”. How “Bonus Seats”, “The 5% limit” and “Rounding Luck” favour some parties more than others. These factors might bend the result in favour or against the UNP + SJB.

But which is it “in favour” or “against”?

In this article, I attempt to answer this question.


Our last general election was in 2015. The UNP came top with 106 seats, while the UPFA got 95.

I consider a hypothetical result: if the UNP contested 2015 on a UNP + SJB split. We assume that the total UNP votes were split equally between the UNP and the SJB. Note, this is an assumption — which might not hold in practice. Also, the equal or 50–50 split also an arbitrary assumption. In practice, the split might be different, especially across districts.

The “Results”

In theory, if the system was completely proportional, there should be no difference. In practice, there is a significant difference.

Data Source: Elections Commission of Sri Lanka

In the UNP + SJB contested 2015 split, the UNP would win 48 seats, and the SJB would win 45 seats. For a total of 93. 13 seats less, compared to no split. The hypothetical split effectively “flips” the election in favour of the UPFA.

[Note: Why did the UNP get three more seats than the SJB, if we split equally? That’s because of rounding errors. In a few cases, the UNP and the SJB need to share an odd number of seats. In these cases, we assigned all the seats to the UNP. In practice, two seat-eligible parties getting the same number of votes is rare.]

What explains the loss? All the factors which we mentioned above.

Bonus Seats (-11 Seats)

In 2015, the UNP got the most number of votes in the Colombo, Gampaha, Kandy, Matale, Nuwara-Eliya, Digamadulla, Trincomalee, Puttalam, Polonnaruwa, Badulla, and the Kegalle Electoral Districts.

On a hypothetical split, they would not get the most number of votes in these districts and hence lose their bonus seat. The alternative district winner (in 2015, mostly the UPFA) would benefit from the UNP’s bonus seat losses.

Rounding Luck (-1 Seats)

On a split, the UNP would lose one seat in each of Matara, Batticaloa, Digamadulla because of rounding effects. However, the UNP would also gain one seat in Trincomalee and the National List.

In general, “Rounding Luck” should mostly “cancel out”.

5% Limit (-1 Seats)

In Jaffna, the UNP got one seat with 6.67% of the vote. On a split, the UNP and SJB would get about 3.33% each, or below the 5% threshold, a party must get to be eligible for seats. Hence, neither would win a single seat.

[UPDATE 7/6] Addendum: The Converse

We see that the UNP + SJB looses about 13 seats by contesting seperately, if they get the same total votes. What about the converse question? How many additional votes must the get to get the same number of seats? That “bump” is around 7%.

Concluding Caveats

This article is not a prediction. It is purely an exercise to demonstrate how our proportional representation system behaves when a party splits.

Also, we considered a 50–50 split. Which is the worst-case scenario, because it maximizes the “Bonus Seats effect”. In practice, the split might be more asymmetric.

With CoViD-19, comic infighting within all parties, and many other nuances, 2020 is likely to be different from all previous elections.

Caveat emptor!




Articles on Politics, and Political Science, by Nuwan I. Senaratna

Recommended from Medium

US Grants Waivers to 8 Countries Including India Importing Iranian Oil: Report

Are Xi and Trump really so different?

The intellectuals of Yemen

A Tale of Two Countries; Let’s Talk Foreign Development

Belarus Is Not Putin’s Ally

How do Guerrilla groups finance their wars?

This election is about the city – not Brexit

BREAKING: Ayade of Cross River sack four of his commissioners, others

BREAKING: Ayade of Cross River sack four of his commissioners, others

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Nuwan I. Senaratna

Nuwan I. Senaratna

I am a Computer Scientist and Musician by training. A writer with interests in Philosophy, Economics, Technology, Politics, Business, the Arts and Fiction.

More from Medium

Sermon by the Rebbitzman  “I read the Torah so you don’t have to.”

Are you a headline voter?

Overpopulation is an utter HOAX. And here’s why: