XTREME Civics & Comparative Government: Voting Systems

Welcome to the first installment of XTREME Civics & Comparative Government. This series is meant to be less specific political issues, but rather to be an explainer of how things work around the world, and to probe at some of the theory around politics as well as challenge some assumptions we all might hold about how things ought to work that may be shaped by the way the society in which you were raised engaged with the public sphere, and less by what is actually working. In the spirit of transparency, I am not a political scientist, and I do have a number of biases. I’ll attempt to minimize these as best I’m able, as well as call myself out from time to time, but please feel free to push back in the comments if you see something. One of the things I want to impress on people reading this: very rarely is there a “right” choice in anything we’re going to look at. There’s simply a choice, and in that choice, a statement of values.

Let us suppose we have a nation, let’s call it, I dunno…the Avocado. For simplicity, let’s say there are five states in the Avocado: Monday, Tuesday, Wednesday, Thursday, and Friday. Each state has a population of 1,000 people, and each state sends 20 Leader Beans to the unicameral Avocado Parliament, because fair’s fair, and they all have the same population. 

The Avocado has two major political parties: the OT Party and the PT Party, bitter ideological rivals. The vast majority of Avocadoans vote for one of these two parties. However, it also has two minor parties, each of whom has a regional base of support. The History Party, whose voters live mostly in Tuesday, and the Magazines Party, who call Friday their home. Voters never behave this logically, but let’s say, for sake of simplicity, that the Avocado’s ideological spectrum runs like this:

So while there are four parties, there are two pairs of ideologically aligned-parties: two parties on the Silly side of the spectrum, and two on the Serious side of the spectrum, with the minor parties the more extreme ones. So a major party voter will always prefer either the other major party or its ideologically aligned minor party over the minor party on the other side of the spectrum from it.

Also for simplicity (and so I don’t have to do lots of allocation and apportionment equations, which will be a future entry), let’s say that while there are 20 separate elections in each state, the states are perfectly uniform districts so that the parties receive the same share of the vote in each election. So we’ll just look at the statewide results, that look like this:

Now, the Avocado has many ways it might have voted, and we’re here to discuss a few of them.

The first system is called a plurality voting system. In many countries, it’s known as First Past The Post (FPTP), and it has a lot to recommend it. For starters, it’s simple and easy to understand. Each voter gets one vote, they vote for the candidate they want to win, and the candidate with the most votes wins. It’s a simple system for deciding who wins in a single election. 

So here’s what the Avocado Parliament would look like under this system:

Because the OT Party won three states, it gets 60 seats, or 60% of Parliament, the History Party gets 20 seats, or 20% of Parliament, and the Magazines Party also gets 20 seats, or 20% of Parliament.

If you’re a voter for the PT Party, you get nothing. You won no elections, so none of your representatives sit in government. 

Even though you’re the second largest party in the country. And in many states, you lost by very narrow margins.

All the other parties are overrepresented: the OT Party has a supermajority in the Parliament despite winning <50% of the popular vote, the History Party has 5 more seats than it should have won, and the Magazines Party has doubled their representation in Parliament over their percentage of the voters. If you’re a PT Party voter, you’re feeling pretty screwed. 

If you’re a History Party voter who lives in Monday or Wednesday, you’re also feeling pretty screwed: you could’ve voted for the PT Party and gotten another Serious candidate in Parliament, but instead you’re stuck with a Silly representative, just because you voted for the person you wanted to win, rather than against the person you least wanted to win. You’re being punished for playing “spoiler” in this election. If you’re a Magazines Party voter in any state but Friday, you also “wasted” your vote on a candidate who didn’t have a chance. In future elections, you may instead choose to vote for one of the two major parties in order to better influence the outcome of the election.

This is one of the flaws of the FPTP: it strongly encourages what’s called “tactical voting,” or voting in a way to maximize the influence of your vote, rather than voting in a way that represents your ideal election outcome. While in many systems voters have come to view tactical voting as a necessary evil, political philosophy generally views it as a bad thing, since voters must naturally suppress their initial voting lean in order to vote in a way that matters. FPTP also encourages gerrymandering: how you draw districts matters greatly. If the states were drawn differently, for instance, you could have a system where only the OT and PT parties won seats by breaking the smaller parties’ votes into other states.

Consequently, FPTP is one of the lesser-used systems in representative democracies, but it’s widely used in countries familiar to the English-speaking world—many countries that were ruled by the British Empire at one point probably use it. That includes the UK, the US, Canada, and India. 

However, because plurality voting has such strengths—voters understand how winners are awarded, there’s no math, they get to actually vote for specific candidates, as opposed to parties—many countries are loath to rid themselves of it. Occasionally, a party that is disadvantaged by FPTP proposes changes, but because to win an election under FPTP things had to break your way under FPTP, very rarely do they keep to that proposal once they take power.

But suppose they didn’t wish to abolish FPTP wholesale, they just wished to undo the vagrancies of the system. One way to do that is to establish a limit: candidates must win elections with at least an absolute majority, or 50% of the vote, plus 1 more vote. 

These systems are called majoritarian systems. The two most notable are two-round elections and Instant Runoff, often conflated with Ranked Choice, Voting (IRV/RCV)—they differ in very specific ways we can’t get into right now. In a two-round system, the top two vote-getters advance to a special runoff election. Because there are only two candidates, it is extremely likely one person will get 50%+1 of the valid votes cast. In IRV and RCV, voters don’t cast a single vote in the general election, but instead rank their choices from, in this case, 1 to 4. 

Because this is an example, Avocadoan voters will behave perfectly rationally and rank in terms of ideological closest match first, then next closest, and then furthest, and no one will stop ranking after a certain number (in some IRV systems, this spoils the ballot and invalidates those votes; in others, it’s acceptable), so whether the Avocado uses a two-round or IRV/RCV system won’t matter, both systems will achieve the same results, which are as follows:

And here’s what the Avocado Parliament looks like based on that result (we’re keeping Tuesday the same because the History Party won an outright majority in the first round):

We’re closer to a system that represents what the voters actually voted for, but we’re also not proportionally representative in a number of new ways. For instance, in the runoff results, the PT Party actually lost votes and the OT Party gained as PT voters in Friday voted to deny the Magazines Party any seats in parliament. Consequently, the Magazines Party voters have no representation at all. The OT Party, despite having won more votes than the PT Party in both the first round and the runoff, is tied with it for seats in parliament. And because the History Party is overrepresented, they hold the balance of power and are likely to reach an agreement with their ideological partners, the PT Party, for control of the government, even though, together, they won just 50% of the vote in the first round, and 45% in the runoff.

One way to handle this is to award seats based on the percentage of the votes received, in a method called proportional voting. In its simplest scheme, you simply take the result of the national election, and award seats. Here’s how the Parliament would look, taking the national result:

Another method would be to award each state’s 20 seats proportionally by the vote in that state, then sum that up, which would look like this (in this case, assigning seats is easy: every 5% = 1 seat, and then the party with the largest remainder is awarded a seat should there be any left over). That would look like this:

As you can see, using states rather than the national results causes each party to be overrepresented or underrepresented in the national parliament, but only by a single seat. And the ideological coalitions retain the same combined percentages in the national parliament that they have under a nationally proportional system. Better than plurality/majoritarian systems, though.

Where it’s not better is local representation. Under pure proportional systems like these, you generally cannot vote for individual candidates. Instead, voters vote for parties, who create lists of candidates who will hold office if the party should win a certain amount of seats. In a closed list proportional system, the party alone decides in what order those candidates will decide who sits in parliament. So, for instance, while the Magazines Party is almost exclusively a Friday party, the national Magazines Party might prefer to seat people who only live from the other four states. 

In an open list proportional system, the party chooses the candidates, but voters are allowed a second vote, in which they indicate (or sometimes rank) which candidate(s) they would prefer to be seated from their chosen party. While this allows individual voters to choose candidates, it’s still incumbent on the parties to put up candidates who will represent the voters.

It’s also not necessarily immune to gerrymandering—proportional systems demand multi-member districts and the number of seats assigned to those districts matter. Using only two-member districts, as in Chile, essentially locks in a two-party system. Three- and four-member districts are better, but can still be drawn in ways to divide a party’s influence. It takes about five or more seats before a district becomes immune to how the boundaries are drawn. 

So suppose the Avocado wanted to keep local representation, but also have a more proportional legislature. One way to get the best of both worlds would be to combine proportional representation systems with plurality/majoritarian voting systems. These are called mixed voting systems, and hang onto your butts, because it’s about to get weird.

In mixed systems, half the seats in parliament would be awarded at the district or constituency level using a plurality or majoritarian voting system. The other would be awarded at a larger level, either regional or national using a proportional voting system. So, in our system, states would now have 10 votes each for 50 locally elected seats, and there would be another 50 seats awarded proportionally based on the national vote.

One of the simplest ways of mixing is parallel voting (also called mixed-member majoritarian or MMM). In this system, voters are given two votes: one for their local FPTP election, and one for the party they’d like to win the proportional election. The results of the state FPTP elections have no bearing on the results of the national proportional election. Instead, proportional seats are divided and then added to the local votes. This system is used in countries like Japan, South Korea, Italy, Russia, and Mexico. 

In our example, a parallel vote might look like this:

Here’s how that calculation would look:

As you can see, we’re closer to proportional than a pure plurality or majoritarian system. Every party has representation. And voters were allowed to pick candidates locally. But where we’ve missed is that the PT Party is still horribly underrepresented, and the OT Party has almost a complete majority. With its ideological partner in the Magazines Party, it has an even larger majority coalition than it possessed as the sole majority party in the FPTP election. In some cases, this might be desirable: because our ideological divides are precisely even, it’s difficult to imagine how a government might form if the legislature had split down the middle. Parallel voting generally encourages more controlling majorities than a system that allows pure proportional representation but is more representative than an FPTP system. It is still susceptible to gerrymandering, but to a lesser degree than a pure FPTP system.

Another way we could hold this election is to award the national seats in such a way that they balanced the allotments of the state victories proportionally. This system is called mixed-member proportional (MMP). It’s also sometimes referred to as a compensatory system because the proportionally awarded seats are “compensation” for parties underperforming in the FPTP elections. 

In MMP, just as in MMM, voters are generally given two votes: one vote for their local candidate, and one for the party. The total result of the regional or national party vote is then used to calculate how many seats a party would be expected to get, and then subtract the results of the local FPTP elections from that number to determine how many proportional seats they should be awarded. It would look like this:

Yes, this is exactly the same image as the nationally pure proportional parliament above. That’s the point of MMP.

And here’s how we got there:

As in our very first example, the PT Party lost every local election and wound up with zero seats. Under MMP, they would be awarded 35 out of the 50 proportional seats because that’s how many they would need to get the 35% of parliament they should have in a proportional system. The OT Party, despite having won 60% of the seats awarded at the local level, are awarded another 10 seats out of the proportional seats to get them to the 40% of the parliament they’re owed. The History Party, also having overshot its mark at FPTP, but undershot when the full 100-member parliament is considered, gets 5 proportional seats. And the Magazines Party, having won exactly the right number of seats it needs, gets 0.

In this way, MMP strikes a balance between proportional and FPTP. Voters can still determine the fates of individual legislators, and parties can still have lists (MMP can also use open or closed lists), but the system will aim to create proportionality. It is also virtually immune to the effects of gerrymandering: because an equal number of seats will be awarded proportionally, drawing districts to benefit a single party also serves to disadvantage them.

What it is not immune from is tactical voting, at least in many places. Because in many systems voters are given two votes, one for a candidate and one for a party, it is possible to split your ballot in such a way to advantage other parties. For instance, OT Party voters in safe OT Party states might choose to give their party vote to the Magazines Party, thereby increasing the proportional seats the Magazines Party would win. If a party were particularly devious (and the law allowed it), it might split into two parties, one that would contest only at the state level and one that would only pursue the regional or national level, thereby turning the MMP system into a parallel voting system where it won more than its fair share of seats in the FPTP elections, and then its partner party was given an overabundance of proportional seats. In countries with a viability threshold to entering parliament, it may also be advantageous to help a party that may form a coalition with your party that is in danger of falling under that threshold by giving them a vote as well.

So there we are: three categories of ways to vote and six different ways to vote. Each has its own drawbacks, each has its own advantages. Which system is how you vote? Which system is how you’d like to vote? What system seems like the worst to you?

P.S. I’ve left out two types of voting systems, one very important, and one not so important, which I’ll cover in a supplement next week. That’s because this was already running long and they’re very peculiar systems that deserve their own attention. Those systems are Single Transferable Vote and the US Presidential Preference Primary systems.