B) "The goal of an election is to give one group the power to rule. Give them a clear mandate to resolve necessary choices." We could call that the dominance purpose of an election. This view risks turning to dictatorship: If the biggest party should dominate a government, should the biggest subgroup control the biggest party? And should the biggest sub-subgroup... This article shows that inclusive decisions can be more stable.
Some people waste whole lifetimes trying to convert good democrats into good authoritarians or vice versa. Both philosophies can work well, for those who like them -- but don't expect authoritarians to build democracies. (Frank Sulloway shows in Born to Rebel . that most firstborns learn to be more authoritarian than their younger siblings.) Ultimately, voting cannot satisfy people with opposing values. Voting with your feet is the surest way to arrive at the policies you want. When that is not practical, build democratic organizations with democrats.
The evolution of democracy speeds up during eras such as The Enlightenment, when many people rejected faith, obedience, and ideology for the hard mental work of rationalism, skepticism and empiricism. Here is one outstanding example.
"No criterion for evaluating voting systems appears
more persuasive than that by the Marquis de Condorcet."
Samuel Merrill III, 1988
Voting is sometimes accused of polarizing electorates. If many voters feel they are placing chips in opposition to other voters, then the community does indeed have a problem. Good rules cannot heal social rifts by themselves, but accurately showing all opinions can move the process forward. Bad rules exacerbate real and contrived conflicts.
Yes-no ballots and plurality rules count chips placed in direct opposition. The rules explained in this article do not, so they do not encourage that way of thinking. These rules may change our fundamental concepts and expectations of voting and government from tools of cultural war between interest groups to tools giving structural support to diversity and its freedoms.
.
What's wrong with the common election rules? For 1 thing, they are inherently erratic in their results and that is closely related to some types of manipulation. Each voter gets 1 vote under the simplest plurality rule, or a number of votes equal to the number of seats to be filled under the bloc vote rule. The candidates that get the most votes win. This usually elects people to represent the plurality or largest group. (Two plurality rules can elect minority reps. Limited vote, gives each voter fewer votes than the number of seats. Cumulative vote, lets each voter give more than 1 vote to the same candidate.) All plurality rules reward finding opportunities to cast a deciding vote, the 1 vote needed to make a loser win, with no votes to spare. Votes that do not help build exactly enough support for a win are called wasted votes. Excess votes for a winner are just as wasteful as votes for a loser. Neither will effect representation on the council.
Gerrymander involves drawing many 1 seat districts where we win with 51% and a few where they win with much more than 51%. That means many of their votes are wasted. Similar tactics work under the bloc, limited, and cumulative vote rules. Encourage rivals to nominate only 1 candidate so she wins with excess votes, all wasted. Multi-winner elections under Proportional Representation make gerrymanders much less effective. These rules put an end to divide and conquer.
Free rides: Don't vote for anyone who has lots of support. She doesn't need yours. Save your votes for candidates where you can make a difference. Of course, when many people take free rides, even the best candidates can lose.
Most point voting rules offer easy strategies:
Punishing votes: Give the minimum to any candidates you don't like, or that rival and threaten your favorite. This is often true even if the rival is your second favorite.
Exaggerations: Give the maximum allowed to the favorite candidate who seems to have a good chance of winning.
The next web page, enact.htm, takes a brief look at manipulations that have more to do with legislative voting. It also explains why STV is the hardest rule to manipulate.
These voting strategies cannot manipulate the Condorcet, STV, or Loring rules explained in the following sections. Most readings in the Resources section at the end of this web page explain other serious defects in plurality rules.
Changing a voting rule may seem a risky step. But STV and Condorcet's rule have been tried and tested for 100 years. Now computers make them easy to tally. The greater risk is continuing to use ancient rules that make democracy work poorly.
.
Preference ballots minimize those negative traits and consequences. They ask voters to rank the candidates as first choice, second choice, third, and so on. Tie votes are allowed. Voters rank only as many as they care to -- but more complete ballots may be more effective. These ballots reveal that the old us versus them political dichotomies actually have many subtle shades of opinion.
Consider a three-way race. A plurality ballot would produce just 3 factions, each supporting 1 candidate. But a preference ballot would reveal 6 preference orders: (1st, 2nd, 3rd) (1, 3, 2) (2, 1, 3) (2, 3, 1) (3, 1, 2) and (3, 2, 1). Tie votes add (1, 1, 1) (1, 1, 3) (1, 3, 1) (3, 1, 1). Such subtle differences do influence who wins by the STV and Condorcet rules. So black and white differences meld into grays, then hues of some color. And more candidates add more colors. Four candidates give 24 preference orders, 6 give 720, and 8 give 40,320.
The cost of voting is a major consideration in creating or choosing a voting rule. Costs may be measured in time, or anxiety.
Some people have thought that preference ballots and their tally rules make ballot counting too hard. But that fear should no longer block the evolution of democracy. Today, the most time-consuming tallies are made easy by small computers and easy-to-use software.
PREFERENCE BALLOT
Candidate Rank Tie Votes Allowed A B C D E F Perot O O O O O O Clinton O O O O O O Dole O O O O O O ? O O O O O O.
PREFERENCE BALLOT
Candidate Rank Tie Votes Allowed Best Worst 1st 2nd 3rd 4th 5th 6th Dole O O O O O O Perot O O O O O O Clinton O O O O O O ? O O O O O O.
.
If a group gets 1/3 of the votes, it should get 1/3 of the seats; that is its fair share and its proportional representation.
Most English-speaking countries elect just 1 person to represent the plurality or largest group in a district. So only the plurality has the right to representation. But most European nations use Proportional Representation (PR). They elect several people to represent a large district. A district with 3 reps might elect 2 from the largest party and 1 from another party. This gives representation to more people and opinions. The more seats in a district with PR, the more accurately its reps speak for its voters.
The plurality rule, on the other hand, exaggerates regional differences. In the North for example, a liberal majority may win all of the seats, while in the South, a liberal minority wins no seats. This may cause regional laws and cultures to diverge and conflict at the national level to increase.
Proportional Representation makes parties campaign for votes everywhere, not just the few, hard-fought swing districts targeted under plurality rule. And they must govern with an eye to voters everywhere. In contrast, a one-winner rule pushes a rep to put constituent service and pork for the narrow interests of her small district above the greater good.
Some writers theorize that negative campaign ads work poorly if a strong third choice is available because the swing voters can turn against those who throw garbage as well as their targets. But in practice, some multi-candidate primaries in the USA have been just as nasty as one-on-one campaigns. Those primaries use plurality rule. Voting rules that make candidates seek 2nd as well as 1st choice votes probably increase the backlash from negative ads because voters committed to the garbage target will decide not to give 2nd choice to the garbage thrower.
Turnout for many US primaries is only 20%. Most voters ignore primary votes. But later, many feel the 2 finalists offer too little choice. Single Transferable Vote combines the primary with the general election: Each party nominates more candidates than it can elect and the voters at the general election decide which nominees are best. In this way, STV increases voters' choices, power, and turnout.
A great advantage of STV is that it needs no parties. That means it can be used by organizations which haven't any such as schools, clubs, unions and so on. And STV voters are never constrained to parties. They may disregard party labels and mix nominees together, ranking a Blue candidate first, an Orange second, another Blue third, a Purple fourth and so on.
the other widely-used form of Proportional Representation is called party-list PR or simply list PR. Ballots for this type of PR only offer voters a choice of parties, not of candidates. Each party is then given a share of seats equal to its share of votes. The first candidate on a list gets the first seat her party wins. Party leaders usually write the lists, so they have immense power over junior politicians and voters. Public, intra-party competition is therefore rare under list PR. This system is used in most European countries.
By contrast, in the STV form of PR, a liberal must compete against similar reps and challengers for the favor of liberal voters. (Sabbatical terms also make elections more competitive, forcing current liberal reps to run against former liberal winners.) This intra-party competition tends to weaken party leaders and increase the power of voters to choose issues and candidates.
To spread power and broaden the appeal of a party's list, its convention (or primaries) could use Loring Ensemble Rule or STV to elect the nominating committee -- whose members take turns adding names to the list.
Under party-list PR, a minority may create a party, succeed at electing reps and yet make its voters less powerful. Its reps might be powerless, left out of the ruling majority. Its voters might seem unobtainable and unimportant to the parties with power. Official policies and ads by all parties then may become more divisive, aggravating social conflicts. But STV allows ballot transfers between parties and limits party discipline, so it isolates groups less than list PR. Loring Ensemble Rule adds a powerful chairperson elected by Condorcet's rule under which all voters are obtainable and needed.
Proportional Representation is most often criticized for creating too many small parties, no majority party, and weak coalition governments. History has shown that the more parties there are in a legislature, the sooner disagreements break the ruling majority apart and force new elections. But a Condorcet legislative rule may find a different majority coalition for each issue. So there is no ruling group and no powerless group. The central chairperson on a LER council also works to create bridges between reps. Reps are elected to set terms of office so governments do not fall.
Political Sim draws a circle around the central voters after each election. The circle encloses only 50% of the simulated voters. But it usually includes 100% of the reps from STV elections for 7 seats or less. (Note the gray circle on the chart.) Moderates on or just inside that statistical line routinely beat fringe candidates.
A candidate must win a quota of weight to win a seat. The "Simple quota" for winning 1 of 3 seats is one-third, or 33%. (There is another type of quota, called Droop quota that adds 1 to the number of seats.) When a candidate gets exactly quota she wins and her supporters' weights are completely used up. They have no more electoral weight, instead they now have a representative. (See STV in Pictures, step 5.) (A voter who did not rank all of the candidates could have all of his choices eliminated before his weight is used up. His weight would be thrown away. The quota for the next winner would be reduced to weights left / seats left.)
If a candidate gets more than enough votes, a share of the excess transfers to each supporter's next choice -- so excess votes are not wasted and every ballot counts equally (step 4). The ballots are counted again. If no candidate gets quota, the weakest candidate (the 1 with the fewest top preferences) is eliminated. Her supporters' ballots transfer to their next preferences -- so votes for losers also are not wasted (steps 2, 3, and 5). Some candidates must lose, but none of the voters should. They will help elect another candidate -- probably one with similar opinions who appeals to more voters.
As long as a top preference has a chance, a rank for a lower preference cannot hurt because it has no weight. But when a top choice has no chance, the ballot can transfer to its next highest choice, if the next choice is known.
This process of eliminations and transfers repeats until all of the voters' weights are used up and all of the council seats are filled. Through transfers, every voter and every opinion group wins its share of representation.
The legislative voting pages explain why STV and LER are the hardest rules to manipulate.
.
An innovation here could change the nature of politics from a war of win or lose to a reasoned negotiation. The decisive swing vote should belong to the center -- not to the left or the right. To achieve that we need to elect about 20% of the council by the most consistently centrist rule: Condorcet's.
Responsive democracy seeks to make the council's views as similar to the voters' as is practical. Accurately matching the median is very important.
LERa is unlikely to elect any other centrist during the STV tally because the Condorcet winner holds the top preferences and weights of some central voters. This rule might be accused of short changing any narrow-minded centrists. It elects strong advocates from other groups; each STV rep works for the interests of the large minority that elected her. But because all voters effect which central candidate wins, she is beholden to and an advocate for the whole.
On the other hand: A) Balanced majorities mean fringe reps rarely influence one-sided majorities, as they often do under PR. With LER they get a voice but little chance to effect policies. B) The centrist is always given the powers of the chair and usually the swing vote. So she has more power than other reps. The trade-off is that other voters help decide who the most central candidate is. C) LERb and LERc do not count the chair as a PR seat for a majority group - so it may get an extra seat. LERa rarely gives the majority an extra seat. Appropriate uses for each ensemble rule are given in the section on Rules, Seats, and Vote Shares.
The biggest winners are the moderates of both sides who always have some power in centered majorities, but who share in it only half the time with one-sided majorities.
(Some LERa simulations end with a fringe candidate and the Condorcet winner - who may not be eliminated. The fringe candidate is eliminated and her supporters' weights transfer to the center. Rare LER or STV simulations force the last weights to transfer even further, from 1 side to the other.)
LER makes elections more competitive and the legislative process more cooperative. The function of legislation is broad cooperation regarding common resources, natural monopolies, and other market failures. The function of elections is competition (mostly between parties) of proposals and competition (mostly within parties) for jobs.
LER supports cultural diversity and thus choices available to individuals. It is open to change via new parties and within parties yet it avoids policy changes that are random or excessive. It reduces the game-of-chance and show-business aspects of politics. This will be an important step in the evolution of stronger democracy.
Major parties or coalitions rarely win over 60% of the PR seats. That must be reduced by 1 out of 6 to become less than 50% and prevent forming an off-center majority. So at least one-sixth of the seats should be filled by Condorcet's rule. Some of those Condorcet seats will be won by non-centrist supporters of the largest party so a larger fraction is better. A 100 seat council might have 80 reps elected by list PR and 20 by Condorcet's rule. A party then needs 1.25% of the votes to win 1 PR seat, 64% for 51 seats.
Voting rules must be adapted to organizational cultures. Small councils have to make tradeoffs, deciding which qualities are most needed: electoral competition, minority representation, or policy moderation. In deciding the voting rule and number of seats for a council, consider the shares of votes for current interest groups. A group with 17% of the population could more likely win 1 of 5 PR seats than 1 of 4. So that district might elect a council of 5 by LERa or STV rather than by LERb. An 83% majority can control the chair by Condorcet's rule anyway. Better yet, expand the council. The framers of a charter or constitution may find simulations of current factional voting and a variety of rules helpful in finding the best rule for their needs.
A unified majority can win control of a council under any fair election rule. Minorities in that case need the protection of the courts. Yet LER can help some: Condorcet's rule gives minority voters some say on which majority candidate becomes the chairperson and STV gives them their own voice on the council. (New voting rules for setting budgets can give minorities limited shares of spending power.)
9-seat councils should include 2 Condorcet Series winners. They will compete to define the best-balanced policies. The 3 or 4 liberals will compete with each other for liberal voters and the 3 or 4 conservatives will do likewise. LERb more often than LERa leads to inter-party competition for a swing PR seat. So for more than 5 seats use LERb or LERc.
7-seat councils are the usual size for "action-taking groups" according to research cited by Mancur Olson in The Logic of Collective Action. These include sub-committees of corporate boards and of state and national legislatures. Larger councils are often used for information gathering.
(Cities or schools can elect a diverse executive council of 5 or 7 even when communication by way of geographic units, barrios or homerooms, requires a large legislative council of single-winner districts.)
5-seat councils are the smallest that can represent a community which has 3 kinds of issues, not only left versus right but a 3D "issue space". (The second dimension is often authoritarian versus democratic.) LERa tends to elect 4 reps near the corners of a 3D pyramid with the chair at their center. She may ally with 1 rep from each side of an issue to make a balanced majority. Five-seat PR districts "waste" fewer votes than 3-seat districts, that means fewer votes go to parties that fail to win a seat and each party's share of seats is much closer to its share of votes.
Clubs, churches, and co-ops rarely have competitive elections. Instead they plead for volunteer directors. It might seem they have little need for better voting rules. But LER or STV can improve representation even when there are only 4 candidates for 3 positions. And after its tally, LER can check whether there is a large sector of voters with no rep. When there is, the by-laws may call for a second election with additional candidates. Or they may call for the Condorcet Series rule which elects the candidates nearest the center and so reduces both the councilŐs inclusiveness as well as its "skewness" off-center.
3-seat councils that use consensus may find STV the best election rule because it might represent a greater number of interest groups and range of opinions than LERa. But when a board votes on some decisions, the 2 rep majority would tend to leave out the third rep's point of view. LERa is better here; its chairperson tries to center and balance each policy. She participates in the discussion, explains the option she prefers and asks the reps to move closer to it. The winner is the one who moves closest to the chair's balanced view. This is not a one-sided victory. It gives only slightly more to 1 side than to the other. If it is a bad compromise the chair may be replaced at the next election.
Germany's 1953 combination of party-list PR and single-winner districts is almost an ensemble rule but: Most single-winner districts are not huge and heterogeneous. They use a plurality rule so reps usually come from and ally with 1 of the 2 large moderate parties not a more central party. The biggest party always forms a ruling majority with a minor party - excluding the second largest party from policy making.
.
"The Choice of Voting Systems", by Richard G. Niemi and William H. Riker; in Scientific American, June 1976.
"Dimensional Analysis of Ranking Data." by Henry E. Brady in American Journal of Political Science. 34 (11/90). Brady found that a spatial model with 2 or 3 opinion dimensions depicted a real electorate fairly well. (Knowing a voter's position on 1 issue helps predict his positions on the related issues in that dimension. Knowing a voter's 3D position helps predict his ranking of candidates.)
Electoral Systems in Comparative Perspective. edited by Joseph F. Zimmerman and Wilma Rule, 1994. The authors look closely at the effects of electoral systems on election of women and minorities.
Making Multi-candidate Elections More Democratic. by Samuel Merrill III, 1988. The main text is for lay readers while math proofs are confined to 6 appendices.
Political Sim. by Robert Loring. Election-simulating software is the best way to learn many voting concepts. A help file within the program expands on this article. Click to see a screen from the game version.
PR Proportional Representation, the Key to Democracy. by George H. Hallet Jr. 1940.
Real Choices / New Voices: the case for Proportional Representation in the United States. by Douglas Amy, 1993.
Representation. edited by Hanna F. Pitkin, 1969. The surprising history and subtle philosophies of representation are reviewed with precision in the editor's 20 page introduction.
Seats and Votes, the Effects and Determinants of Electoral Systems. by Rein Taagepera, and Mathew Soberg Shugart; 1989. The authors combine the work of many others to build a mathematical model of political systems.
"Social Choice Observed: Five Presidential Elections of the American Psychological Association" by John R. Chamberlin, Jerry. L. Cohen, and Clyde H. Coombs in Journal of Politics. 46 (1984): 479-502. The authors used computer simulations to find the Condorcet efficiencies, violations of subset rationality, and frequencies of manipulable elections for 5 voting rules.
Topics in the Theory of Voting. by Phillip D. Straffin, 1980. Straffin's short book is the clearest introduction to voting rules. He analyzes the faults possible under each rule. He does not write about the frequencies of failure as Merrill did.
The Tyranny of the Majority, by Lani Guinier, 1994.
Voting and Democracy Report: 1995. edited and published by The Center for Voting and Democracy. Authors from Lani Guinier to Clarence Thomas argue for using PR in the USA.
.
Return to Democracy Evolves!
Move on to enacting policies.
Go to site outline.
© 1997 Robert Loring Reprints permitted
