It amazes me how little demand there is for truly fair democracy. Even though government too often ignores the will of the people, and even though many go to the polls knowing that neither of the two frontrunner candidates have a platform they like, they never complain about the systems that select these candidates.
In the long run, the only way to have better government is to have better governmental systems. And the cornerstone of any democratic system is its election methods. The single-plurality (SP, also called first-past-the-post) systems used in Canada, the USA and elsewhere are poor systems which tend to reduce the diversity of candidates, opinions, and political platforms.
Doesn't it bother you that in most elections, people must choose between two dominant candidates, or two dominant parties? People are far too diverse for a mere two candidates or parties to represent everyone effectively. Doesn't it bother you that voting for a non-dominant party effectively wastes your vote, because when you choose a third party, you give up the chance to decide which of the top two you would prefer? Haven't you ever voted for one of the top two, even though you disagree with the candidate's platform in several areas? Haven't you ever cast a vote for a candidate you didn't particularly like, simply because "the other guy was worse"? Hasn't it ever bugged you when all the major candidates completely ignore an issue that really matters to you?
Have you ever witnessed a vote split spoil an SP election? Here's an example. Suppose 60% of an electorate are conservative, and 40% liberal. It stands to reason, then, that the candidate elected should be somewhat conservative too. (And, of course, this discussion would apply equally well when we swap the words 'conservative' and 'liberal'; election methods are a nonpartisan issue!). Now, suppose a third party or independent conservative candidate gains substantial support among the people--let's suppose most conservatives even prefer this third party to the incumbent. There are many reasons why this could be; sometimes it is because the third candidate is especially charismatic and has a platform that strikes a chord with the people; other times it is because the people are discontent with the incumbent conservative. In these situations, the conservative voters are in a conundrum. They know that the third party is unlikely to win, because of the simple fact that it is a third party. On the other hand, they do not like the incumbent conservative, but most would prefer not to vote for the liberal. In a case like this, a vote split usually occurs. Some of the conservatives vote for the third party, either as a 'protest' vote, or because they actually believe the third party has a chance of winning. Meanwhile, the other conservatives decide they would be throwing their vote away if they were to vote for the third party, so they do not. Therefore, most vote for the disliked incumbent, and a few vote for the liberal, on the basis that anything's better than that crappy incumbent.
A typical result, then, might be that 45% vote liberal, 40% for the incumbent, and 15% for the third party. Thus, despite a strong conservative majority, the liberal wins. Further, even though the third party is the most preferred, it gets the least number of votes. I think it's important to note how the availability of more than two choices leads to perverse results. Because of this, the vote-splitting problem, diversity of candidates is actually undesirable in a SP system. A party never offers multiple candidates in an SP general election, because it would hurt or eliminate the possibility that any of its candidates would be elected. And, again because of the VS problem, it is rare for more than two parties or candidates to have a chance of winning (at least in any single electoral district, or 'riding' as it is called in Canada.) In summary, voters have only two realistic choices most of the time, and should a third party get a significant number of votes, the outcome is often perverse.
Finally, note that results like the example above are usually transient. Voters, after seeing the third party's poor results, typically decide it won't be possible after all to elect the third party, and they return to voting for the incumbent. On rare occassions, the third party might actually win the subsequent election, but only if the two dominant candidates are particularly abysmal. Note well that the third party cannot win on his merits alone; he must also be so much better than the other candidates that voters are willing to risk a vote split to get him into office. Generally this means that not only must the third party be really good, but both of the other candidates must be really bad. In other words, third party contenders face a huge barrier to entry. But this barrier need not be so huge.
I wonder how people so easily accept the perverse properties of single-plurality systems. Are they simply ignorant of the fact that there are superior alternatives? Do they think that the system used to select a nation's leaders doesn't really matter? Or do they simply think that there's no way the present system can change, and that therefore there is no reason to discuss or promote better systems?
I admit to not having a thorough knowledge of all election systems; my main purpose in writing this is not to pose myself as an expert on the topic, but to remind people that we can do better. I do know of two good alternatives to SP for choosing one candidate from many, approval voting and Condorcet voting, which I learned about at electionmethods.org. Unfortunately, the site's webmaster has decided to abandon democracy and remove all the information there, because a colleague of his believes in some Bush conspiracy theories. Frankly, it's a damn poor reason.
This is a very simple voting method that produces good results. It mostly solves the VS problem, but not completely. It is just like single plurality, except that voters can choose more than one candidate. Thus, voters are not faced with the dillema of throwing their vote away if they would like to vote for a third party.
A variation on Approval voting is cardinal voting, where you can specify a "percentage" or a number on a scale, indicating "how much" support you give to each candidate. However, Condorcet voting (below) is more mathematically sound, and the ballot is no more complicated than Cardinal voting. The reason I would support Approval voting is that its simplicity makes it easier for people to support, since, I worry, many people aren't willing to use their brains well enough to understand Condorcet. However, I figure, if people are willing to think their way through cardinal voting, they should be willing to think their way through Condorcet also. In that case, there's no reason to favor cardinal over Condorcet.
Mathematically, this is the best way to choose a single person from many candidates. Condorcet is a preferential ballet system based on the principle of pairwise races--comparing every candidate to every other candidate. Condorcet allows every voter to completely express his preferences, and then these preferences--all of them--are used in pairwise races to determine the winner. For any two candidates A & B, the voter can state one of three preferences:
(a) "I would rather elect A than B"
(b) "I would rather elect B than A"
(c) "A and B are equal to me. In a race between A and B, I
don't care who wins."
Occasionally when Condorcet is adopted, choice (c) is not allowed, but I can't imagine a good reason for that.
After the ballots are collected, every pair of candidates is compared. Typically, there is one candidate that beats every other candidate in every pairwise race. In that case, that candidate wins. Sometimes, three or more candidates can beat each other in a "circle" or "cycle" of pairwise races; for example A beats B, and B beats C, but C beats A. There is more than one way to deal with such situations, but basically it involves looking at by how much each of the candidates beat each other.
How might a Condorcet ballot look? One approach is a series of numbers, or check boxes, beside each name, where number one represents the highest preference, and higher numbers represent lower preferences:
| Preference | : | Highest ... Lowest |
| George W. Bush | : | 1 · 2 · 3 · 4 · 5 |
| John Kerry | : | 1 · 2 · 3 · 4 · 5 |
| Ralph Nader | : | 1 · 2 · 3 · 4 · 5 |
| Etc... | : | 1 · 2 · 3 · 4 · 5 |
| Etc... | : | 1 · 2 · 3 · 4 · 5 |
You would then circle the numeral (or check the box) representing your preference for each candidate. If, for a given candidate, you do not circle a number, it would normally indicate the lowest possible preference, meaning you would rather elect anyone else.
It's important to understand that it doesn't matter what number you select for a candidate in isolation; only the relative number matters. First of all, it should be clear that if one circles "1" for all candidates, or "5" for all candidates, one may as well not vote at all; the ballot has no effect. Now, suppose there are three candidates, A, B, and C. A ballot that selects 1, 2, 3 (for A, B, and C, respectively) has exactly the same effect as a ballot that selects 2, 3, 4. Likewise, a ballot selecting 1, 4, 5 also has the same effect. Clearly, there's no reason to have more ranks than candidates, and for most voters in most situations, I expect 5 ranks should be sufficient to describe all the candidates in which the voter is interested. Finally, a vote for candidate A has no effect on the race between B and C. For example, if I rank Ralph Nader as my first choice, it has no effect on the race between Bush and Kerry.
In theory, the ballot could contain another option, which I will call ND (No Decision). If you circle ND, it indicates you don't want to specify any preferences for that candidate with respect to the others. The ND option indicates that you choose option (c) in all pairwise races involving that candidate. The effect of NP is to allow all other voters (who did not choose ND) to decide the result of the election for that candidate. To put it another way, NP means "I haven't made a decision about this candidate, so I will let everyone else decide." This would be a useful option in case of ignorance: you don't know anything about the candidate, or you have heard some good things and some bad things, and you honestly can't figure out if you want the person in office. However, I doubt an option like this would ever be present on a real ballot, because the people would start complaining that it's too complicated and that they don't understand it. Personally, I think I would use an option like ND on occassion, because I often don't know enough about candidate to decide. A ballot could even offer a "range" of indecision, but perhaps I'm the only voter that would use that. Experience makes me think that most people who are too ignorant to make a decision are not willing to admit it, not even to themselves, while filling out a secret ballot.
You may have heard of another voting system called IRV--Instant-Runoff Voting. IRV has none of the mathematical elegance of Condorcet, but for some reason I do not comprehend, it is more well known and more popular. It is actually used in Australia, but sadly it has not improved the amount of choice in elections very much. IRV is also a preferential voting system, and the actual ballots of Condorcet and IRV are essentially the same, except that IRV does not allow choice (c), equal preference. Thus, provided (c) is not allowed, the ballots can be identical.
There are other interesting systems out there, such as proportional representation (PR), but I don't know enough others yet to discuss them here.
It would be very difficult to (peacefully) change the systems we use now, but it is not impossible.
I don't know the specifics of the systems of most countries, but for any country, it would at least require that the majority of the legislative body vote in favor of changing the system, which would generally require a party, or coalition of parties, that support electoral reform to be elected. Invariably, the parties now in power do not support electoral reform because it would weaken their power; therefore, either a third party would somehow need to rise to power, or else a major party would have to be 'iniltrated' somehow by people that support electoral reform.
In any case, I believe a prerequisite for electoral reform would be that the majority of the populace support it; such support, in turn, would require that the people become educated about it. One way to accomplish this would be to find some niche, some "district" in the country which
Once one or several such small districts begin to use Condorcet, I speculate that the media could be made to take notice, which would allow the education process to begin. If Condorcet can be demonstrated to have substantial benefits, and if people can be educated about these benefits, and if the two ruling parties have made the people particularly discontent with their poor leadership, the people might be willing to elect a third party in favor of electoral reform--if not on a national level, then at least on a provincial (or U.S. state) level. I see a catch, though: people would not elect the third party on the basis of electoral reform alone. A good platform beyond that, a platform broadly supported by the people, would be required.
Electoral reform is not a magic bullet. It is a basis for further change, based on greater democracy, on putting power in the hands of the people. Democracy is not perfect, but the alternative is worse and can be seen everywhere in "democracies" today. The alternative is to put most of the power in the hands of a few only--the rich, the elite, the executives of corporations, and the big stakeholders in the same, who care about money and power more than all the people they walk on to get it.