# Electoral Systems: Other Ways to Skin the Cat

Electoral Systems: Other Ways to Skin the Cat

The nation will soon be voting, and voters will soon be wondering whether their votes have really been counted for the candidates they favor. This wake-up call reminds us that the election system is a tool for achieving a certain purpose and that it doesn't necessarily work as well as it might.

The election system doesn't include merely the mechanical tools we use (punch-ballots, voting machines, etc.). It includes also the arrangement of choices we make as voters. For example - to give just a superficial listing -- do we cast one vote for either of two candidates? One vote for either of several candidates? Several votes among a number of candidates? (I'm talking only about voting for officeholders, not about voting for or against ballot issues.)

We usually don't even notice the system of choices we engage in, much less the alternatives, because we take the present system for granted - just as we took the mechanical tools for granted until Florida in 2000.

But if we look more deeply, and more widely, we'll see that there are other ways, perhaps better ways, of achieving the purposes of voting.

These purposes are: 1) To register the wishes of the voters and translate their individual wishes into one overall decision; 2) In so doing, to give all voters an equal voice.

Let's review the obvious: The present system of voting allows a voter to cast one vote for one of a number of candidates (usually there are only two feasible candidates); or if there are a number of positions to be filled, to cast as many votes as there are positions.

This present system may seem obvious and natural - and perhaps it was when the nation was founded. But at that time the electorate was much more homogeneous than now: it was completely male, propertied, white, and almost entirely Western European (especially British) in origin. Things have changed.

The present system is subject to at least two basic flaws:

VOTER FUTILITY. In some cases, a number of voters really have no effect, because they are constantly and inevitably outvoted. In effect, they are disenfranchised. This is most obvious in the case of deep bias against identifiable minorities, such as blacks or Hispanics, but it is also true of districts in which one party is a permanent majority (gerrymandered districts furnishing the most prominent examples).

In some cases, of course, the minority group might gain power by negotiating with one part or another of the majority group - the minority's support in exchange for concessions from the majority. However, this won't hold true when there is an unchangeable conflict between the minority and majority, as in the case of racial animosity (or between Republicans and Democrats, nowadays).

DIVIDED COMMITMENT (too many candidates). Suppose there are three candidates, A, B and C, trying for one seat, and suppose that A and B share the same viewpoint. Suppose that 60% of the voters favor that viewpoint, while only 40% of the voters favor C's viewpoint. But suppose further that those voters who favor A-and-B divide their votes equally between A and B. In that case (barring a runoff), C would win with 40% against 30% for each of the others.

Of course, primaries avoid this sort of situation, but primaries have their own deep faults; as we have seen so well, they most strongly attract the extremists from each party, while automatically excluding those who don't identify with either party. Runoffs also have their difficulties.

In general, the present system of voting runs up against a dilemma. Let's call an election arrangement "tight" when there is only one more candidate than office(s) to be filled (e.g., two candidates for one office). Let's call it "loose" when there are more then one extra candidate (e.g., three candidates for one office, or seven candidates for three). Then "tight" arrangements are subject to Voter Futility, and "loose" arrangements are subject to Divided Commitment.

The Great Decisions group at Lakewood Library took up the subject of alternative voting systems at a September meeting. We found the subject a daunting one, requiring close examination, but well worth the effort. Three alternative systems (in addition to interest voting) came to our attention. The first was Balanced Voting (or Negative Voting), championed by Lakewoodite Calvin Wight.

Balanced (Negative) Voting

The Balanced Voting plan is quite simple: Each voter is given one positive vote (as now) plus one negative vote. Thus a voter may vote in any of three ways: for a candidate; against a candidate; or for one candidate and against another. The votes against a given candidate are subtracted from the votes for, and the candidate with the best net score wins.

This of course won't work in a two-candidate race. The results would be the same as under present arrangements because every voter would vote for one candidate and against the other.

But look at what would happen in a three-candidate race. Take the example given above, in which the A-B position commands the allegiance of 60% of the voters, while the C position has only 40%. Recall that under the present system, candidate C, representing a minority viewpoint, would probably win.

But with Balanced Voting, the A-B voters could all vote against C, giving him a net score of minus 20% (40% for and 60% against). Meanwhile, the C voters could all vote against A and B. If these votes were distributed evenly, A and B would each get a negative score of 20% and would wind up with a net score of plus 10% (30% for and 20% against). So either A or B would win the election, and the majority A-B position would carry the day.

Thus Balanced Voting would solve the Divided Commitment problem while making primaries unnecessary and allowing all voters to vote for all the candidates. It would also spur voter examination of the candidates, since voters would often want to know not only whom to vote for, but also whom to vote against. It would obviously allow for a finer discrimination among candidates, as voters selected one candidate among several to vote against. Candidates in addition might try to be less divisive than otherwise, in order to avoid the negative votes from those with differing viewpoints. Finally, the plan is simple and therefore easy to understand.

Proportional Representation

As the name implies, proportional representation aims to give all the various groups in the electorate the number of representatives proportionate to their strength within the community, thus solving the Voter Futility problem. For example, a group with 40% of the voters should get 40% of the representation, instead of being completely locked out as could happen in the present system.

The first requirement for proportional representation is to be rid of single-seat districts. For example, if a legislature has 50 seats, these could be divided into 10 districts with five representatives apiece, instead of 50 districts with one representative apiece. Or all representatives would represent the community at large, as would be appropriate to the first of the proportional representation plans, called Party List Voting.

Party List Voting

Throughout the world, this is the most common type of proportional representation. Under this system, voters vote primarily for a party, not individuals. Each party is awarded a percentage of the representatives equal to the percentage of the vote it gets. For example, suppose party A gets 40% of the vote, party B gets 35%, and party C gets 25%. Then seats are awarded accordingly: 40% for A, 35% for B; 25% for C. The candidates on each list would have been ranked, and the top candidates, to a number equal to their allotted percentage, would be picked to serve. (E.g., in this example, if there are 100 total seats, the top 40 candidates from party A would be elected, and so on.)

How are the lists ranked? There are two ways:

Closed list voting: The ranking of the candidates is done by the party (perhaps through a primary). The voters are simply presented with the ranked lists and know that the top people on the list will be first to be chosen.

Open list voting: The voters themselves do the ranking at the same time they vote for the party. In other words, the voters vote for a party (as described above) and at the same time they vote for a candidate within that party. The results of this second vote determine how the candidates are ranked.

Single Transferable Vote Method (Choice Voting)

This method of proportional representation is the most fun, though also the most complicated. It was employed in Cleveland, precariously, from 1923 to 1931, and it is used now in Cambridge, Massachusetts (where, I understand, it is computerized, a real help as you will see).

As the name implies, this method allows for transfer of votes, collecting all the votes for one point of view under one or a few candidates. Thus if the candidate you favor does not get elected, your vote is not wasted because it is counted for someone else who represents your viewpoint.

In this method, voters don't just vote for a person or persons. Rather, they write down a number in front of each candidate they care to vote for, expressing their preferences in rank order. The write "1" beside their first choice, "2" beside their second choice, and so on.

Then the votes are counted. The first step is to establish a number called the "Threshold" or "Quota." This is the number of votes a candidate needs to be elected. The formula is as follows:

Threshold = (number of votes/number of seats + 1) +1

This is better grasped by example: Suppose there were 120 votes cast and 3 seats to fill. The number of votes (120) is divided by 4 (the number of seats plus one), equaling 30. Add one to that number and you get the threshold number, which is 31. If enough candidates to fill the seats (namely 3) got this number of votes, then no other candidate could get more than they. (If three candidates got 31 votes apiece, the total for the three would be 93, and only 27 votes would remain for any other candidate.)

After the Threshold number is established, the ballots are examined to see which candidates, if any, have reached the threshold on the strength of their first-place votes. These candidates are of course elected. Afterwards, the process consists of lopping off the winners and the losers (those in last place) and giving their votes to the candidates listed next on their ballots, proceeding in this way until enough candidates have been elected to fill the required number of seats. The process, in essence, goes like this:

For a WINNING candidate (if any), all votes in excess of the Threshold are transferred to the next choice (i.e., the candidate marked as #2.)

If this results in a sufficient number of candidates making the Threshold, the election is over. Otherwise, the candidate with the LEAST number of votes is eliminated, and his/her votes are transferred to the next choices on his/her ballots.

If this results in a sufficient number of candidates making the Threshold, the election is over. Otherwise, the procedure is repeated until a sufficient number of candidates are elected.

Interest Voting

One consideration is neglected in all the methods I have touched on, namely, the intensity of voter interest. Some voters passionately desire their favorite to win, while others still favor the candidate but with only a moderate degree of interest. Yet both types of voter are accorded only one vote. The passionate voter could complain that he is not being accorded equal treatment when his passionate interest has no more effect than the feeble interest of someone else.

To be sure, under the present system a voter may express interest in a candidate in ways that go beyond merely voting, namely by working for a candidate or contributing money. But some voters are far more able to give their time than others, and the disparity is even greater when it comes to contributing money. So in effect we are authorizing inequality if we count on campaign work or financial contributions to express the voters' different degrees of interest.

It is possible, however, to incorporate expression of interest into the voting system itself, by allowing the vote to indicate level of interest. For example, a "3" stands for great interest, a "2" for moderate interest, and a "1" for slight interest. The numbers for each candidate would then be added up, of course, and the candidate with the largest number would win.

This runs the danger that voters will give the highest number to every candidate they favor, thus defeating the purpose of allowing for different levels of interest. This danger can be overcome by putting a cap on the total number a voter can register. For example, suppose the cap was 12, and suppose there were four candidates. Then votes of "4," "3," "3" and "2" would be legitimate, whereas "4," "4," "4," "4" would not be.

The nation will soon be voting, and voters will soon be wondering whether their votes have really been counted for the candidates they favor. This wake-up call reminds us that the election system is a tool for achieving a certain purpose and that it doesn't necessarily work as well as it might.

The election system doesn't include merely the mechanical tools we use (punch-ballots, voting machines, etc.). It includes also the arrangement of choices we make as voters. For example - to give just a superficial listing -- do we cast one vote for either of two candidates? One vote for either of several candidates? Several votes among a number of candidates? (I'm talking only about voting for officeholders, not about voting for or against ballot issues.)

We usually don't even notice the system of choices we engage in, much less the alternatives, because we take the present system for granted - just as we took the mechanical tools for granted until Florida in 2000.

But if we look more deeply, and more widely, we'll see that there are other ways, perhaps better ways, of achieving the purposes of voting.

These purposes are: 1) To register the wishes of the voters and translate their individual wishes into one overall decision; 2) In so doing, to give all voters an equal voice.

Let's review the obvious: The present system of voting allows a voter to cast one vote for one of a number of candidates (usually there are only two feasible candidates); or if there are a number of positions to be filled, to cast as many votes as there are positions.

This present system may seem obvious and natural - and perhaps it was when the nation was founded. But at that time the electorate was much more homogeneous than now: it was completely male, propertied, white, and almost entirely Western European (especially British) in origin. Things have changed.

The present system is subject to at least two basic flaws:

VOTER FUTILITY. In some cases, a number of voters really have no effect, because they are constantly and inevitably outvoted. In effect, they are disenfranchised. This is most obvious in the case of deep bias against identifiable minorities, such as blacks or Hispanics, but it is also true of districts in which one party is a permanent majority (gerrymandered districts furnishing the most prominent examples).

In some cases, of course, the minority group might gain power by negotiating with one part or another of the majority group - the minority's support in exchange for concessions from the majority. However, this won't hold true when there is an unchangeable conflict between the minority and majority, as in the case of racial animosity (or between Republicans and Democrats, nowadays).

DIVIDED COMMITMENT (too many candidates). Suppose there are three candidates, A, B and C, trying for one seat, and suppose that A and B share the same viewpoint. Suppose that 60% of the voters favor that viewpoint, while only 40% of the voters favor C's viewpoint. But suppose further that those voters who favor A-and-B divide their votes equally between A and B. In that case (barring a runoff), C would win with 40% against 30% for each of the others.

Of course, primaries avoid this sort of situation, but primaries have their own deep faults; as we have seen so well, they most strongly attract the extremists from each party, while automatically excluding those who don't identify with either party. Runoffs also have their difficulties.

In general, the present system of voting runs up against a dilemma. Let's call an election arrangement "tight" when there is only one more candidate than office(s) to be filled (e.g., two candidates for one office). Let's call it "loose" when there are more then one extra candidate (e.g., three candidates for one office, or seven candidates for three). Then "tight" arrangements are subject to Voter Futility, and "loose" arrangements are subject to Divided Commitment.

The Great Decisions group at Lakewood Library took up the subject of alternative voting systems at a September meeting. We found the subject a daunting one, requiring close examination, but well worth the effort. Three alternative systems (in addition to interest voting) came to our attention. The first was Balanced Voting (or Negative Voting), championed by Lakewoodite Calvin Wight.

Balanced (Negative) Voting

The Balanced Voting plan is quite simple: Each voter is given one positive vote (as now) plus one negative vote. Thus a voter may vote in any of three ways: for a candidate; against a candidate; or for one candidate and against another. The votes against a given candidate are subtracted from the votes for, and the candidate with the best net score wins.

This of course won't work in a two-candidate race. The results would be the same as under present arrangements because every voter would vote for one candidate and against the other.

But look at what would happen in a three-candidate race. Take the example given above, in which the A-B position commands the allegiance of 60% of the voters, while the C position has only 40%. Recall that under the present system, candidate C, representing a minority viewpoint, would probably win.

But with Balanced Voting, the A-B voters could all vote against C, giving him a net score of minus 20% (40% for and 60% against). Meanwhile, the C voters could all vote against A and B. If these votes were distributed evenly, A and B would each get a negative score of 20% and would wind up with a net score of plus 10% (30% for and 20% against). So either A or B would win the election, and the majority A-B position would carry the day.

Thus Balanced Voting would solve the Divided Commitment problem while making primaries unnecessary and allowing all voters to vote for all the candidates. It would also spur voter examination of the candidates, since voters would often want to know not only whom to vote for, but also whom to vote against. It would obviously allow for a finer discrimination among candidates, as voters selected one candidate among several to vote against. Candidates in addition might try to be less divisive than otherwise, in order to avoid the negative votes from those with differing viewpoints. Finally, the plan is simple and therefore easy to understand.

Proportional Representation

As the name implies, proportional representation aims to give all the various groups in the electorate the number of representatives proportionate to their strength within the community, thus solving the Voter Futility problem. For example, a group with 40% of the voters should get 40% of the representation, instead of being completely locked out as could happen in the present system.

The first requirement for proportional representation is to be rid of single-seat districts. For example, if a legislature has 50 seats, these could be divided into 10 districts with five representatives apiece, instead of 50 districts with one representative apiece. Or all representatives would represent the community at large, as would be appropriate to the first of the proportional representation plans, called Party List Voting.

Party List Voting

Throughout the world, this is the most common type of proportional representation. Under this system, voters vote primarily for a party, not individuals. Each party is awarded a percentage of the representatives equal to the percentage of the vote it gets. For example, suppose party A gets 40% of the vote, party B gets 35%, and party C gets 25%. Then seats are awarded accordingly: 40% for A, 35% for B; 25% for C. The candidates on each list would have been ranked, and the top candidates, to a number equal to their allotted percentage, would be picked to serve. (E.g., in this example, if there are 100 total seats, the top 40 candidates from party A would be elected, and so on.)

How are the lists ranked? There are two ways:

Closed list voting: The ranking of the candidates is done by the party (perhaps through a primary). The voters are simply presented with the ranked lists and know that the top people on the list will be first to be chosen.

Open list voting: The voters themselves do the ranking at the same time they vote for the party. In other words, the voters vote for a party (as described above) and at the same time they vote for a candidate within that party. The results of this second vote determine how the candidates are ranked.

Single Transferable Vote Method (Choice Voting)

This method of proportional representation is the most fun, though also the most complicated. It was employed in Cleveland, precariously, from 1923 to 1931, and it is used now in Cambridge, Massachusetts (where, I understand, it is computerized, a real help as you will see).

As the name implies, this method allows for transfer of votes, collecting all the votes for one point of view under one or a few candidates. Thus if the candidate you favor does not get elected, your vote is not wasted because it is counted for someone else who represents your viewpoint.

In this method, voters don't just vote for a person or persons. Rather, they write down a number in front of each candidate they care to vote for, expressing their preferences in rank order. The write "1" beside their first choice, "2" beside their second choice, and so on.

Then the votes are counted. The first step is to establish a number called the "Threshold" or "Quota." This is the number of votes a candidate needs to be elected. The formula is as follows:

Threshold = (number of votes/number of seats + 1) +1

This is better grasped by example: Suppose there were 120 votes cast and 3 seats to fill. The number of votes (120) is divided by 4 (the number of seats plus one), equaling 30. Add one to that number and you get the threshold number, which is 31. If enough candidates to fill the seats (namely 3) got this number of votes, then no other candidate could get more than they. (If three candidates got 31 votes apiece, the total for the three would be 93, and only 27 votes would remain for any other candidate.)

After the Threshold number is established, the ballots are examined to see which candidates, if any, have reached the threshold on the strength of their first-place votes. These candidates are of course elected. Afterwards, the process consists of lopping off the winners and the losers (those in last place) and giving their votes to the candidates listed next on their ballots, proceeding in this way until enough candidates have been elected to fill the required number of seats. The process, in essence, goes like this:

For a WINNING candidate (if any), all votes in excess of the Threshold are transferred to the next choice (i.e., the candidate marked as #2.)

If this results in a sufficient number of candidates making the Threshold, the election is over. Otherwise, the candidate with the LEAST number of votes is eliminated, and his/her votes are transferred to the next choices on his/her ballots.

If this results in a sufficient number of candidates making the Threshold, the election is over. Otherwise, the procedure is repeated until a sufficient number of candidates are elected.

Interest Voting

One consideration is neglected in all the methods I have touched on, namely, the intensity of voter interest. Some voters passionately desire their favorite to win, while others still favor the candidate but with only a moderate degree of interest. Yet both types of voter are accorded only one vote. The passionate voter could complain that he is not being accorded equal treatment when his passionate interest has no more effect than the feeble interest of someone else.

To be sure, under the present system a voter may express interest in a candidate in ways that go beyond merely voting, namely by working for a candidate or contributing money. But some voters are far more able to give their time than others, and the disparity is even greater when it comes to contributing money. So in effect we are authorizing inequality if we count on campaign work or financial contributions to express the voters' different degrees of interest.

It is possible, however, to incorporate expression of interest into the voting system itself, by allowing the vote to indicate level of interest. For example, a "3" stands for great interest, a "2" for moderate interest, and a "1" for slight interest. The numbers for each candidate would then be added up, of course, and the candidate with the largest number would win.

This runs the danger that voters will give the highest number to every candidate they favor, thus defeating the purpose of allowing for different levels of interest. This danger can be overcome by putting a cap on the total number a voter can register. For example, suppose the cap was 12, and suppose there were four candidates. Then votes of "4," "3," "3" and "2" would be legitimate, whereas "4," "4," "4," "4" would not be.

Volume 2, Issue 22, Posted 2:02 PM, 10.26.06