The New Electoral Map of Sri Lanka (Part 1)
EMMA and Assumptions
This is the first of 4 articles where we discuss constructing a new Electoral Map for Sri Lanka.
In this article, we discuss how to apply EMMA (Electoral Map Mapping Algorithm) to Sri Lanka, and what assumptions we must make.
In How to Draw Sri Lanka’s New Electoral Map, we discussed the workings of an Electoral Map Mapping Algorithm (EMMA).
As we saw, EMMA takes the Old Electoral Map of Sri Lanka as input. It modifies it step-by-step, gradually making it more and more optimal in terms of the five properties proposed in “5 Properties of Good Electoral Maps”. Once “sufficiently optimal”, EMMA outputs the New Electoral Map.
In this article, we will look at EMMA’s output.
Assumptions
Before we proceed, let’s look at two assumptions that we’ve made.
Electoral Map with 160 Polling Divisions each with one seat
EMMA generates an Electoral Map dividing Sri Lanka into 160 Polling Divisions, each with one seat. Since EMMA respects Electoral District (ED) boundaries, these 160 seats are first allocated amongst the 22 Electoral Districts.
The seats allocated by EMMA to each ED differs from the 2020 (and other) Parliamentary Election Allocations because EMMA allocates only 160 seats to the EDs, while in Parliamentary Elections, 196 seats are allocated (of which 160 are allocated proportionally and 36 based on province, somewhat disproportionally).
Population vs Registered Voters
EMMA allocates seats proportional to population, while in Parliamentary Elections, the 160 seats allocated proportionally are allocated in proportion to the number of registered voters.
EMMA uses population (from the 2012 census) because it needs population data at the Grama Niladhari Division (GND) level. As of now, we don’t have access to registered voters at this level. Hence, EMMA cannot use registered voters for proportional allocation.
Compared to the Parliamentary Elections allocation, EMMA’s allocation is likely to favour populations a smaller proportion of registered voters — i.e. those with younger populations.
Assumptions out of the way, we will now discuss EMMA’s results grouped by the strategies it uses to modify the Old Electoral Map (How to Draw Sri Lanka’s New Electoral Map describes these strategies in detail).
In Part 2, we construct electoral maps for 12 Electoral Districts, which require Merging, Splitting existing polling divisions, or a combinations of Merging and Splitting.
In Part 3, we construct electoral maps for the remaining 10 Electoral Districts, which need combinations of Merging, Splitting, or Merging and Splitting.
In Part 4, we analyze our Electoral Map relative to the “5 Properties of Good Electoral Maps”.