First things first: Venezuela’s National Assembly has just one chamber. The National Assembly to be chosen on December 6th, 2015 will have 167 diputados (MPs). 164 of them will represent given areas: states or circuitos. The last three are representatives of indigenous Venezuelans. Parliamentary elections are held every five years.
The voting system is usually called a “mixed” system because some seats are elected by majority rule constituencies and others through a list system meant to ensure proportional representation. But, as we will see, neither part of that description is quite right. In reality, it’s a strongly majoritarian election system. In most circumstances. Be warned, generalizations can be tricky.
Let’s start with constituencies – known locally as circuitos. 113 diputados are elected by a simple majority rule (a.k.a., “first past the post”) in districts all around the country. Most of them are single-member constituencies, which simply put means the candidate with most votes on his name is elected, but there are exceptions (because easy and straightforward would be unacceptable in Venezuela). For example, Miranda State’s Circuit number 6 chooses two diputados, but they still run separately. In effect, these are “first two past the post” constituencies, where the two top vote-getters get seats in the A.N., whether or not they represent different parties.
Why does CNE do it this way instead of breaking big circuitos up into two? Nobody knows.
Then we have the list vote. The rationale behind having lists in addition to circuits is to guarantee an element of proportional representation. List seats are apportioned using the dreaded D’Hondt Method. The D’Hondt Method is the electoral equivalent of fútbol’s off-side rule – obvious to devotees, totally obscure to casual watchers. (You can find a detailed description here, if you dare). But the essence is this: each state gets a number of diputados to be divided proportionally between parties according to their result statewide.
Once upon a time, this proportionality element was applied on a national scale and resulted in actual correspondence between the percent of the vote a party got and its number of seats in parliament. Neither of those things are true anymore.
Why? Because in all but three states there are only two seats to be allocated to the lists! In practice, that means that in virtually all cases, the top two most popular parties in a state will get one seat each.
An example: Imagine Party A gets 40% of the votes, Party B gets 30% Party C gets 25% well Party A would get 50% of the seats and Party B would get the other 50%, leaving Party C and 25% of the population without representation. Doesn’t sound so ‘proportional’ now, does it?
So in fact, we have 113 seats decided on a first past the post (or first two past the post) basis, three that go to indigenous representatives, and more or less half of the remaining 51 that end up divided fiti-fiti in hands of the top two parties. The result is a system that’s “proportional” only not really. When you fiddle with simulations, you realize the system is strongly majoritarian, as “first past the post” systems tend to be – except when the election is close.
What the system does do is over-represent rural states, which pick more representatives per head of population than urban states. This is a consequence of having only one chamber, because you must design the system in a way that small states are guaranteed some weight in the House. Historically, the opposition vote has been bunched mostly on urban areas, while chavismo has been stronger on rural states; giving them an additional edge. As a consequence simulations tend to favor PSUV if the election is close, but switch to creating oversized majorities for whichever side is the winner, when the election is not close.
And this election seems anything but close.